Download The job creation potential of the service sector in

Employment Observatory
RESEARCH Network
The job creation potential of
the service sector in Europe
Final report 2000
Edited by
Dominique Anxo and Donald Storrie
(Centre for European Labour Market Studies, Sweden)
In collaboration with the following research group members:
Randall W. Eberts, W.E Upjohn Institute for Employment Research, Kalamazoo, USA
Bernard Gazier, Université de Paris I, Paris, France
John Philpott, Employment Policy Institute, London, the United Kingdom
Günther Schmid, Wissenschaftszentrum für Sozialforschung, Berlin, Germany
Leonello Tronti, Instituto Nazionale di Statistica, Rome, Italy
Françoise Benhamou, Université de Paris I, Paris, France
George Erickcek, W.E Upjohn Institute for Employment Research, Kalamazoo, USA
Dietmar Dathe, Wissenschaftszentrum fur Sozialforschung, Berlin, Germany
Colette Fagan, Department of Sociology, University of Manchester, the United Kingdom
Roberta Sestini, University of Rome, ‘Tor Vergata’, Rome, Italy
Nadine Thevenot¸ Université de Paris I, Paris, France
Andrea Toma, Centro Studi Investimenti Sociali, Rome, Italy
Employment & social affairs
Employment and European Social Fund
European Commission
Directorate-General for Employment and Social Affairs
Unit EMPL/A.1
Manuscript completed in 2000
A great deal of additional information on the European Union is available on the Internet.
It can be accessed through the Europa server (http://europa.eu.int).
Cataloguing data can be found at the end of this publication.
Luxembourg: Office for Official Publications of the European Communities, 2001
ISBN 92-894-0603-8
© European Communities, 2001
Printed in Spain
PRINTED ON WHITE CHLORINE-FREE PAPER
Contents
Executive summary: Conclusions and policy recommendations
Dominique Anxo and Donald Storrie
9
9
Objective and structure of the report
9
Main findings
The job creation potential of the service sector: General considerations
The industry and service nexus: Concepts and empirical evidence
New perspectives of the service economy
10
10
14
17
Policy recommendations
19
Introduction
Dominique Anxo and Donald Storrie
25
25
Part I: The job creation potential of the service sector:
General considerations
33
1. Service employment, productivity and growth
Donald Storrie
35
35
1.1. Introduction
35
1.2. The concept of services
37
1.3. Employment in services
1.3.1. Introduction
1.3.2. The long-run growth of aggregate service employment share
38
38
39
3
The job creation potential of the service sector in Europe
1.3.3. The long-term growth of service employment share by subsector
41
1.3.4. Recent growth of service employment rates
46
1.3.5. Concluding remarks on the EU–US service gap by share and rate
51
1.4. Services and economic development
52
1.4.1. Long-term shifts in employment structure
52
1.4.2. Productivity in services
54
1.4.4. Services as a source of economic growth
61
1.5. Conclusions
2. Unbalanced growth and employment in services
62
65
Leonello Tronti, Roberta Sestini and Andrea Toma
65
2.1. Introduction: the Baumol model
65
2.2. The predictions of the model
68
2.3. Relaxing the assumptions and discussing the model
69
2.3.1. Constant labour pool
71
2.3.2. Long-term invariance of sector shares
74
2.3.3. The hypothesis of a small open economy
78
2.3.4. Productivity, wages and technical progress
81
2.4. Looking for alternatives
3. Service employment: A gender perspective
86
89
Dominique Anxo and Colette Fagan
89
3.1. Introduction
89
3.2. The macro societal perspective
90
3.3. The implications of sex-segregated employment, service sector restructuring
and new work organisations for women’s labour market integration
98
3.4. The gender division of labour, the service economy and equal opportunities:
A micro perspective
103
3.4.1. Gender and time allocation in the household
104
3.4.2. The gender division of domestic labour, market substitutes and equal
opportunities
105
3.4.3. The structure of wages, taxes and benefits and the effect
on the consumption of market services
106
3.4.4. Women’s employment and the consumption of market services
107
3.4.5. The job creation possibilities of expanded market provision
of household-related services
107
3.5. Policy implications
4
109
Contents
4. Labour costs, social habits and employment regimes. Two comparisons
of national employment performances and some lessons
Françoise Benhamou and Bernard Gazier
4.1. Introduction
4.2. Employment in the retail trade sector in France and the United States:
An appraisal of the Gadrey/Piketty controversy
4.2.1. Setting the problem: Evaluating the potential job creation of France
in the retail trade sector
4.2.2. The divergences
4.2.3. Assessing the controversy
4.3. Employment in the performing arts and audio-visual activities in France
and the United Kingdom
4.3.1. Employment in cultural activities and Baumol’s cost disease
4.3.2. The growth in cultural employment: Some convergence between France
and the United Kingdom
4.3.3. Toward an interpretation of the divergences between France and
the United Kingdom: A specific way of managing risk
113
113
113
114
114
115
117
120
120
121
126
4.4. Job creation and flexibility in cultural activities: Some hypotheses
129
4.5. Conclusions and policy recommendations
131
Appendix
132
Part II: The industry and service nexus: Concepts and empirical evidence
135
1. The nature and determinants of service sector growth in the United States
Randall Eberts and George Erickcek
137
137
1.1. Introduction
137
1.2. Employment growth trends in the service sector in the European Union
and the United States
1.2.1. Comparison analysis
1.2.2. Employment in the United States service-producing sectors
138
138
140
1.3. The nature of services and its implications for growth
143
1.4. Structure of employment growth: Shifts in occupational demand
1.4.1. Occupational shift-share analysis
1.4.2. The outsourcing of non-core support services in the manufacturing
sector
1.4.3. The use of temporary services
147
148
1.5. Sources of service employment growth
1.5.1. Internal growth in services
1.5.2. The high income elasticity of demand for consumer services
152
152
153
150
150
5
The job creation potential of the service sector in Europe
1.5.3. The lack of productivity growth in services
1.5.4. Employment dynamics
154
156
1.6. Empirical evidence of the manufacturing and/or services nexus
1.6.1. Geographic characteristics of service employment growth
1.6.2. Factors determining service sector expansion
1.6.3. Determinants of service sector job creation
158
158
159
160
1.7. Wage structure
1.7.1. Wage distribution
1.7.2. Age-earnings profile
161
161
163
1.8. Growth potential of a service sector economy
164
1.9. Conclusion
166
Appendix
169
2. Determinants of business and personal services: Evidence from
the German regions
Dietmar Dathe and Gunther Schmid
181
181
2.1. Introduction
181
2.2. Theoretical notes and observations on the service gap
2.2.1. The service gap I: The case of personal services in Germany
2.2.2. The service gap II: The case of business services in Germany
183
184
186
2.3. Structure and dynamics of service employment in German regions
2.3.1. General trends of employment
2.3.2. The influence of sectoral branches on regional employment
2.3.3. Structural and locational determinants of service employment
2.3.4. Employment rates and sectoral mix
192
192
195
196
197
2.4. Summary discussion of the results and policy conclusions
2.4.1. Evidence on business and personal services from Germany
2.4.2. Curing the service gap by wages or consumer subsidies
2.4.3. Evaluation of the present discussion
2.4.4. Policy recommendations
199
199
201
204
206
Appendix
208
3. Analysing business services employment — some theoretical
and methodological remarks
Bernard Gazier and Nadine Thevenot
239
239
6
3.1. Introduction
239
3.2. An optimistic management approach: The ‘service profit chain’
241
Contents
3.3. Combining a labour law and economic approach: Risk management,
risk spillover and the service employment relationship
3.4. An economic approach: The determining factors of business
service externalisation
3.4.1. The heterogeneity of business services: What are the origins?
3.4.2. Business service activities and the ‘make or buy’ dilemma
3.4.3. Proposals for a classification of the forms of externalisation
of service activities
244
248
248
249
253
3.5. Measurement problems and differentiated job creation potential of business
service externalisation
254
3.5.1. Measures of business service externalisation and their limits
254
3.5.2. Organisation of service activities and employment relations:
Some hypotheses about measuring job creation potential
of business service externalisation
256
3.6. Conclusion and policy implications
257
Part III: New perspectives of the service economy
265
1. Weightlessness and the political economy of service sector job creation
John Philpott
267
267
1.1. Introduction
267
1.2. Weightless growth and employment
1.2.1. The knowledge economy
1.2.2. Knowledge as an engine of growth
1.2.3. Explaining the ‘productivity paradox’
1.2.4. Weightlessness and employment
1.2.5. The new micro-service economy
1.2.6. Outsourcing and networking
1.2.7. Households
1.2.8. Weightless worries
1.2.9. New economy, third way
269
270
270
271
273
274
274
276
278
280
1.3. Employment policy in the weightless economy
1.3.1. General conditions for service sector job creation
1.3.2. The relative merits of private sector, public sector and social
economy jobs
1.3.3. Targeting of service sector jobs
1.3.4. The cost of job creation
1.3.5. Productivity, pay and conditions
1.3.6. The scope for job creation in the private, public and social economy
sectors
281
281
283
284
284
284
286
7
The job creation potential of the service sector in Europe
2. Beyond conventional service economics — utility services, service product
chains and job services
Gunther Schmid
2.1. Introduction
287
2.2. From producer services to user services
288
2.3. Repealing cost disease through service product chains
293
3.4. From personal transfers to job services
296
3.5. Conclusion
300
Bibliography
8
287
287
303
Executive summary
Conclusions and policy recommendations
Dominique Anxo and Donald Storrie (1)
Objective and structure of the report
Exploiting the opportunities for job creation in the service sector appears as a major issue in
the European economic and employment strategy. In the 2000 Employment Guidelines, the
European Commission stressed the need to develop a policy framework in order to fully
exploit the employment potential of the service sector. Focus is placed on identifying and
removing those obstacles that continue to hamper the growth of enterprises and the creation
of more and better jobs. The guidelines also emphasise the importance of exploiting the
employment potential of the social economy and information society.
The main objective of this research project is to understand the theoretical basis of the
dynamic of employment growth in services. Apart from the abovementioned policy
perspective, it is clear from the long-term trends in employment structure that net job creation
in advanced economies occurs almost exclusively in the service sector. A further motive for
this study is the observed slow pace of total recent employment growth in the EU compared
to the robust employment growth in the United States. Even though both the EU and United
States have experienced significant growth in service producing jobs during this time, the
share of service employment in the EU still lags behind that of the United States. Another
concern is the growth potential of an economy that increasingly relies upon services.
The following report is structured around three main themes. Part One, ‘The job creation
potential of the service sector: General considerations’, addresses some empirical, theoretical
(1) The executive summary is based on the various contributions. We thank all the authors for constructive
comments.
9
The job creation potential of the service sector in Europe
and gender issues. Part Two, ‘The industry and service nexus: Concepts and empirical
evidence’, focuses on the interaction between manufacturing industry and producer services
in the dynamics of job creation. Finally, Part Three, ‘New perspectives of the service
economy’, reflects upon the impact of the new information and communication technology
for our conception of services and the potential for job creation resulting from advances of this
technology. Furthermore, the job creation potential of services is examined in the light of the
possible switch from product markets to utility markets and innovative possibilities in
managing specific income risks related to modern services.
The executive summary follows the overall structure of the report.
Main findings
The job creation potential of the service sector: General considerations
In Chapter 1, ‘Service employment, productivity and growth’, D. Storrie, introduces the
empirical rationale of the report. Discussion on the potential of job creation in the service
sector in the EU often focuses on the higher growth of the aggregate employment rate and the
larger size of the service sector in the United States. However, while one can also observe a
higher growth of the United States’ service employment rate, exactly half of the widening of
the aggregate EU–US employment rate gap since the mid 1980s is due to a more rapid fall in
European agricultural and industrial employment. Thus, in relative terms, the most striking
difference in the change of employment structure in the United States and the EU during this
period, has been the continuation of the long-term decline in agriculture in the EU, and
particularly in the less developed EU economies.
Thus, compared with the United States, the poor EU employment record is across all three
major economic sectors. Indeed, it is important to note that relatively speaking, European
services have performed well. This is not to say that the basic premise of this report, that there
is a potential for job creation in the service sector, is incorrect. Job creation will almost
certainly continue to be largely in services. However, comparisons with other sectors of the
economy, and in particular some consideration of the European experience of the
simultaneous decline in industry and agriculture with an increase in services, should focus
analysis and policy on the process of structural change.
While one should always be wary of predicting future developments by a simple extrapolation
of current trends, many factors suggest that employment in agriculture (and to a lesser extent
industry) will continue to decline and that it will tend to be re-allocated in the service sector.
In the more developed nations the shedding of labour out of agriculture and industry to
services is a very long-term trend and there is little reason to suppose that it will be reversed.
In the less developed States (in both Europe and elsewhere), the long-term trend has been
somewhat different in that the largest employment sector has shifted directly from agriculture
to services without an intervening period of industrial dominance. There are reasons to
suppose that shedding of agricultural labour from the less developed Member States will
10
Conclusions and policy recommendations
accelerate in the future. The key issue here is, of course, the development of the common
agricultural policy. Despite acute political difficulties, the pressure for reform, that is, a
lowering of protection, and thus employment, is intense.
There has been a continuous process of EU catch-up to United States levels of service
employment share. While the catch-up has generally been faster in the less developed Member
States, it is still these economies that account for most of the EU’s service share gap with the
United States. This further underlines the relevance of viewing the potential for service job
creation in Europe in terms of structural shifts from, primarily, agriculture to services.
Producer services are the source of the largest aggregate EU gap to the United States,
particularly in the less developed Member States. This sector includes many of the jobs that
are typically associated with the rise of new technology and work organisation. While, in this
respect, the EU deficit in this sector may be viewed with concern, it is noteworthy that the EU
catch-up with the United States in producer services is proceeding at a faster pace than in all
service subsectors (both in terms of employment share and rate).
The second largest gap is to be found in personal services. This can mainly be ascribed to the
larger hotel and restaurant sector in the United States. While the gap is present in every
Member State except Spain, the gap is considerably smaller in the less developed Member
States, that is, the opposite of the European distribution in producer services. A smaller gap,
but with the same European distribution, is found in trade and retail. The European gap in
these sectors should not be viewed with the same degree of concern as for producer services,
as one would be less likely to find the type of productivity enhancing spillovers to other
sectors that could be expected in many producer services.
In the largest service sector, social services, the EU average is at the same level as the United
States figure. However, there is a great variation within the EU, particularly in health and
social services. While job quality in this sector is rather heterogeneous, the European shortfall
in health and education may be viewed as problematic, given that growth theory suggests that
these sectors may be very conducive to economic growth, particularly among less developed
economies.
The chapter proceeds with an introduction to an issue taken up in several of the following
chapters, namely that of productivity in the services. The major, and somewhat worrying point
made here is that the difficulties in obtaining empirical measurements of service productivity
are very serious indeed. On the other hand, the possibility that service productivity has been
seriously underestimated may serve to cast doubt upon the gloomy predictions of the Baumol
model presented in Chapter 2. The possibility that service productivity is higher than has been
reported, especially in the light of the new information and communication technology, is also
pursued in both chapters of Part 3.
In Chapter 2, L. Tronti, R. Sestini and A. Toma, ‘Unbalanced growth and employment in
services’, present the most common theoretical model used to analyse the long-term
development of the service sector, namely, Baumol’s model of ‘unbalanced growth’. The key
11
The job creation potential of the service sector in Europe
assumption of the model is that the economy can be divided into two sectors: the progressive
sector with high productivity growth and a non-progressive sector with lower productivity
growth. Parts of the service sector are assumed to be in the non-progressive sector.
By successively relaxing the productivity and other assumptions of the model, the authors
explore some theoretical and policy consequences. The first important result is that service
employment growth should not in itself be an appropriate policy objective unless such growth
takes place in industries that contribute, directly or indirectly, to aggregate productivity
growth. Another implication of the Baumol model is that a major way to escape the ‘Baumol
trap’, or at least to soften its constraints, could be provided by new technologies and
organisation improvements that transfer technical progress into previously non-progressive
service industries. These changes can expand the boundaries of the progressive sector and thus
delay the asymptotic stagnation of aggregate productivity, even though this may leave the
trade-off between productivity and employment growth unchanged.
Relaxing the Baumol assumption of a closed economy, the authors note that the exposure of
international trade may spur productivity growth in the industries concerned and, second, that
sheltered services may profoundly affect the performance of exposed industries if they
provide them with domestically produced inputs. For this reason, a policy aiming at favouring
international competition in the services and increasing domestic competition in the service
industries whose output is complementary to exposed industries, should be promoted. Finally,
the choice of favouring employment growth in the services by differentiating wages along
with productivity growth may prove unsuccessful in terms of welfare, as it may accelerate
productivity asymptotic stagnation even in the presence of some positive effects on aggregate
output growth. Furthermore, in a context of unbalanced productivity growth, the possible
benefits of wage differentiation are, according to the authors, only of a short-term nature.
As roughly 80 % of all female employment in the European Union is to be found in the service
sector, it is obvious that any analysis of the service sector must focus on its gender dimension.
In Chapter 3, ‘Service employment: A gender perspective’, D. Anxo and C. Fagan show that
the increased feminisation of the labour force during the last decades has been closely
associated with the development of the welfare state in general and the growth of services in
particular. Despite the significant reduction of the gender employment gap, there are still large
differences in the patterns of female integration in Europe. However, national differences in
women’s employment rates cannot simply be ascribed to country variation in the magnitude
of the service sector, and the diversity in employment structure appears not to be a good
predictor of female integration in the labour market. On the other hand, the authors show that
the nature of the welfare regime influences women’s employment prospects directly in the
sense that a relatively large welfare state regime increases the demand for female labour and
that countries with a relatively low employment rate tend to have relatively underdeveloped
social welfare services and/or a low share of private service work.
The segregated division of labour within the household around the traditional ‘male
breadwinner’ model of men as wage earners and women as homemakers is being replaced by
specialisation and segregation within the labour market. Gender segregation in employment
12
Conclusions and policy recommendations
remains persistently high despite the feminisation of employment. Women are overrepresented in services connected with providing physical and emotional care and personal
services. As the prevailing situation in most advanced economies is characterised by a high
level of gender segregation and as one cannot expect this to change in the short run, then
further reduction of the gender employment gap will be essentially related to the growth of
female-dominated sectors. Future employment growth is predicted to be concentrated in
service activities and many of these jobs will be connected with welfare services as well as
personal and household related services. How this expansion in welfare services will be
funded, and whether it will be organised through direct public sector provision or through
combinations of public, private and voluntary sector partnership remains unclear. If the
outcome is a reduction in public service provision driven by public expenditure constraints to
meet the convergence criteria of monetary union, one cannot exclude that women will be
affected not only through a potential deterioration in both quantity and quality of public sector
employment, but also because if the public provision of child care and elderly care is cut back,
this will constrain their labour supply.
It has been argued that one benefit of a policy aiming at favouring an expansion of household
related market services may be to promote equal opportunities by allowing women to reduce
their domestic workload by purchasing market substitutes thereby enabling them to compete
with men on more equal terms in the labour market. Another related benefit is that the female
employment rate could be expected to rise, as a large proportion of the workforce in these
household related market services is likely to be women. However, the quality of the jobs that
emerge will also have an impact on the situation of women in the labour market. If the jobs
are low paid or precarious, this will contribute to the developing polarisation between highly
qualified, well-paid women and those who are less qualified. Even if the expansion of market
provision of household related services is organised around better quality jobs, the authors
argue that it will do little to challenge the existing division of household labour between
women and men, given the slow adaptation of men to involvement in domestic work. To
promote equal opportunities, other complementary policies are needed alongside subsidised
household related services, targeted at increasing men’s involvement in domestic labour.
In Chapter 4, ‘Labour costs, social habits and employment regimes’, F. Benhamou and B.
Gazier focus on the importance of employment regimes in explaining the employment gap in
retail between France and the United States and the divergent patterns of employment
adjustment in the French and British cultural sector. The most common and recurrent
explanation for the higher US intensity of personal services is related to the greater dispersion
of the wage structure. The authors contrast this common view with the role of social and
cultural habits. After reviewing both explanations, the authors show that neither are fully
convincing. For example, the labour cost argument fails to explain why a persisting costlowering policy in France did not bring any employment growth in the sector. On the other
hand the ‘societal habits ‘ perspective cannot take into account the rapid employment response
of the retail sector to wage subsidies to promote part-time employment. By introducing the
concept of employment regimes, the competition structure and employment patterns of the
retail sector, combined with the incidence of public employment policies, are shown to be
useful elements in understanding the divergent experiences of the two countries.
13
The job creation potential of the service sector in Europe
The growth of cultural employment in France and the United Kingdom is also a good
illustration of the importance of the institutional framework for explaining national
divergences in the form of employment growth in service activities. During the last decade,
employment growth in the British and French cultural sectors has been impressive. However,
while employment increased at a similar rate, two different models of labour market
adjustment were adopted. In the United Kingdom, there was a strong growth of selfemployment, while in France employment growth took the form of a significant increase in
intermittent employment. These divergent responses appear to be shaped by the specific
national feature of labour market regulation and welfare state regimes. In the United
Kingdom, the general legal and institutional environment, and in particular, the Enterprise
Allowance Scheme, introduced in 1983, boosted self-employment; while in France, the
specific unemployment benefit system for the performing arts and the film industry favoured
the development of precarious work. According to the authors, one cannot conclude in favour
of one model or the other in terms of labour market performance. However, the French model
is probably threatened by the high cost of its unemployment system, while, in the United
Kingdom, deregulation brings about risks borne exclusively by the self-employed.
The industry and service nexus: Concepts and empirical evidence
The second part of the report, ‘The industry and service nexus: Concepts and Empirical
Evidence’, focuses on both the determinants of employment growth in services, in particular
in business services, and on the interaction of service sector and the rest of the economy, in
particular with manufacturing. The empirical basis of the first two chapters relies on regional
data to analyse the manufacturing/service nexus.
The chapter of R. Eberts and G. Erickcek, ‘The nature and determinants of service sector
growth in the United States’, addresses the following basic issues: the determinants of service
sector employment, the relationship between services and manufacturing, the wage structure
of service jobs, and the potential for the service sector to stimulate overall employment
growth. The main focus of the analysis is on business services in the United States, as they
have achieved remarkable employment growth during the past several decades. Business
services are also considered because they are closely related to the restructuring in the
manufacturing sector and provide intermediate inputs to manufacturers. Consequently, factors
contributing to this tremendous employment gain include the outsourcing of non-core
activities. Not only are United States manufacturers losing jobs in many service related
occupations to the service sector, but also the business and professional sectors are gaining
from manufacturers in several production-related occupations. These findings suggest that
temporary and lease employment agencies are increasingly being called upon to meet the
needs of manufacturers. Still, while evidence indicates that structural change in manufacturing
is a contributing factor to the growth in service employment, its impact is small, relative to the
growth in the demand for services by other services.
The lack of productivity growth in services accounts for a significant portion of the gap in
employment growth between services and manufacturing. In fact, since growth in service
output and manufacturing output are roughly the same, the difference in employment growth
14
Conclusions and policy recommendations
between the two sectors is almost entirely due to the difference in productivity growth. This
raises the issue of the ability of services to support innovative activities that are needed to
sustain future productivity growth. It may be the case that while service sector productivity
lags behind manufacturing productivity, innovations in manufacturing are increasingly
dependent upon service activities, such as engineering and research and development. When
these functions were performed internally by firms, the lag in productivity went unnoticed.
However, as these services are outsourced and sold in the market, the tendency to measure
productivity gains in these sectors, however crude and inaccurate, has raised concerns about
lagging service sector productivity growth. However, researchers have suggested that service
sector productivity should be measured by how effective the services are in increasing the
productivity of those activities that use these services as inputs. This approach relates closely
to the standards typically used when these services are provided internally.
Although the lack of productivity growth is a contributing factor to the employment growth
in services, wages and wage growth in services are becoming more comparable to those in
manufacturing. Less-educated workers still find better paying jobs in manufacturing.
However, on average, service workers can expect the same wage growth over their work life
as manufacturing workers.
Observing variations in service and manufacturing employment across metropolitan areas
offers some insights into the factors that contribute to their growth. The United States’
experience testifies that service employment is more volatile than manufacturing employment
and remains concentrated in large metropolitan areas. Furthermore, the regional analysis
suggests that employment in the services expands in regions that have a large and dynamic
manufacturing base, low manufacturing wage rates and a relatively highly educated labour
force.
Finally, the service sector is becoming an increasingly important component of the economic
base of many regions in the United States. Moreover, their research shows that manufacturing
and service activities are complementary. Services are no longer limited to a ‘non-base’
function that simply recirculates revenues brought into the region through the sale of
manufacturing goods. The regional export of services, such as health, financial and
professional services, can have an impact on the region’s economic base, similar to that of
filling a new order in manufacturing. Hence, regional economic development efforts may
become more robust by creating a healthy economic environment for both its manufacturing
and service sectors.
The contribution of D. Dathe and G. Schmid, ‘The determinants of business and personal
services, evidence from the German regions’, identifies the factors explaining the contrasting
employment growth in personal and producer services in Germany and, like the previous
chapter, explores the relationship between employment in the manufacturing and the service
sector by using regional data.
As far as business services are concerned, the empirical analysis supports the concept of the
interactive nature of the knowledge-intensive sector within business services and knowledge-
15
The job creation potential of the service sector in Europe
intensive manufacturing industries. Whereas the overall nexus between industries and
services declines, this nexus is the crucial determinant for a positive and sustainable
employment growth. In line with the results of the previous chapter, the manufacturing
industry remains crucial, but industrial products are increasingly more service intensive. A
corollary feature is the strong correlation between the skill level of the regional labour force
and regional employment growth, especially related to expert-oriented services and business
services.
Regarding the regional dimension, that is, the spatial distribution of activities, the authors
found that high skill, expert-oriented services and knowledge-intensive industries are still
concentrated in metropolitan areas. Low skill services have a stronger propensity to locate
either in the centre of agglomerated areas or at the rural periphery. On the other hand, largescale internal labour markets in manufacturing, located so far mostly in metropolitan areas,
are transformed into network labour markets. The employment complementarity between
knowledge-intensive services and industries spills over into other domains. Increasing
demand in service-intensive manufacturing goods leads to a parallel development of
employment in industries and services; this creates higher income and thereby higher demand
for personal services thus creating a virtuous circle. Interactive services may also reverse the
conventional product cycle: originally intermediate services turn out to become (exportable)
service goods such as software packages, management and expert systems. Exporting services
still require personal interaction, that is, personal presence on the (foreign) markets, thereby
requiring language and communication capabilities. The culture of mobility is vital for
competitive advantages in high skill services. This is a further explanation for the strong
correlation between skill level and expert-oriented services.
D. Dathe and G. Schmid also argue that one of the most important determinants of successful
adjustment to globalisation are information and communication networks. They are the
crucial levers to enhance productivity in business services and knowledge-intensive industries
with likely spillover to personal services. The reason is that information networks increase
their efficiency with rising numbers of participants, probably exponentially. The existence and
public support of such networks explains regional differences. Overall, the results justify the
extension of the ‘industrial district’ to the ‘service industry district’ hypothesis.
As far as personal services are concerned, D. Dathe and G. Schmid confirm the result found
previously in Chapter 3, Part 1, namely, that the service society provides a path for women
into the system of gainful labour market work. This changes the form in which the female
labour potential is organised but hardly the content. Thus, regions with high shares of personal
or social services have a higher female labour force participation and vice versa. They also
find that demand for personal services rises with qualifications, especially with the skill level
of women. Thus, we find higher service employment rates in regions with high skill levels, a
pattern that correlates with agglomeration since the skill and income level in these regions is
higher than in rural areas.
Previous empirical evidence suggests that manufacturers are outsourcing more and more of
their ‘non-core’ functions. At the micro level the theoretical basis of the analysis of whether a
16
Conclusions and policy recommendations
firm performs the service in-house or purchases it on the market is to be found in the theories
concerning the boundaries and structure of the firm. The principle focus has been on
transaction costs, economies of scale and scope and general problems of coordination in
situations of uncertainty. The purpose of the chapter by B. Gazier and N. Thevenot,
‘Analysing business services employment — some theoretical and methodological remarks’,
is to explore some dimensions and determinants of the growth of business services. The
authors begin by analysing business services from a purely managerial point of view, using
the service profit chain approach. They stress, in particular, the importance of trust and loyalty
for business services, where the creation of long-term relationships between the employees
and the customers is mirrored by a correspondingly long-term relationship between the
employees and the employer. In order to secure an efficient service relation, the firm must
provide a degree of work autonomy to secure and motivate employees. The employee’s
loyalty reinforces the client’s trust and loyalty, and thus promotes profitability. However, it is
shown that such a virtuous circle depends on a precise assignment of suitable employees to
particular clients and the degree of service and or product differentiation.
Business services and outsourcing are further analysed within the transaction cost/risk
management framework. At a theoretical level, the authors show that externalisation of
business service activities is intimately related to the level of transaction costs and in
particular the degree of risk related to the activities. The authors propose a classification of
business service externalisation according to two criteria: the possible participation of the user
firm during the conception process and the possible involvement of the user firm during the
production process. These levels of participation depend on the features of services: the
tangibility level of the service support, and the specificity level of the needs of the user firm.
Finally, some hypotheses are developed for the job creation potential of business service
externalisation. First, the job creation potential depends on the potential of obtaining
productivity gains by externalisation. Second, the job creation potential depends on the way
services are conceived and produced. The potential can be assumed higher in the business
services corresponding to high quality jobs for at least two reasons. First, for these services,
the externalisation process does not correspond to a simple transfer with a substitution of
internal and external solutions. Because these business services are co-conceived or coproduced, the growth of the service sector goes with the growth of employment in the user
sectors. Moreover, the relational dimension of these services implies little economies of scale
potentialities. Second, there is a growth of the needs for these services, which influence the
competitive advantage of the firms in a ‘service economy’ (2). Therefore, the job creation
potential would stem as much from externalisation as from the growth of the needs.
New perspectives of the service economy
The final part of this report, ‘New perspectives of the service economy’, takes up some
innovative perspectives of the service economy. As indicated in the 2000 Employment
Guidelines, information technology plays a vital role in the growth of service employment.
(2) See also Philpott in this volume with the ‘weightless economy’.
17
The job creation potential of the service sector in Europe
The new information and communication technology has already had dramatic effects on
various service sectors. This technology could be seen as being the first major technological
innovation that has directly addressed parts of the service sector, as many services often entail
the processing of information. The last two chapters of the report explore the implications of
the new ‘informational paradigm’ and the emergence of ‘utility markets and service product
chains’ on consumer behaviour, work organisation, skill requirement, job creation potential
and productivity growth.
In his chapter, ‘Weightlessness and the political economy of services sector job creation’, J.
Philpott examines both the general conditions for job creation in the new economy and the
criteria by which to judge the relative merits of job creation in the private, public and social
economy sectors. The weightless economy is defined as an economy in which fewer workers
are engaged in the production of tangible goods; those that remain are increasingly
knowledge workers, needed to improve the intangible qualities of products rather than to do
unskilled muscle work or routine skilled or unskilled manual operations. This leaves the
service sector, which produces intangibles, as the major source of net employment growth. In
contrast to the pessimistic accounts of deindustrialisation, predictions of the ‘end of work’ and
the extreme cost disease model, Philpott offers a more optimistic scenario. Human creativity
and skill is presented as being vital to making the most of knowledge-based technology, while
the ‘human touch’ is essential for the provision of a wide variety of personalised services, both
high-tech and low-tech.
J. Philpott ascribes much importance to new technology, and in particular information and
communications, which includes the Internet and media technology. A key characteristic of
the knowledge economy is that the marginal cost of production falls towards zero. As
competition tends to press prices towards marginal costs, this implies that the securing of
intellectual property rights is of utmost importance for the firm’s decision to invest. It also
leads to an increasing importance of advertising and the creation of brand names.
As regards the low reported level of productivity, the author notes both the measurement
problems mentioned above, but appears to ascribe more importance to the slow diffusion of
productivity enhancement due to cultural lags in both producer and consumer behaviour.
Indeed much emphasis is placed on problems related to the diffusion of knowledge. While so
far the high technology, new producer services have largely been outsourced, or rather have
been incorporated into strategic and creative networks, J. Philpott questions whether this will
emerge to be the dominant corporate form. The reason for this is that investors may come to
place even greater weight on the value of companies’ human capital and knowledge talent, and
thus encourage firms to protect knowledge assets by developing them in-house. Likewise,
companies will seek to secure a comparative advantage in product markets by developing tacit
knowledge within corporate boundaries rather than networking on the basis of transfer of
codified knowledge.
The negative side of the knowledge-based economy is not only limited to the job loss in
manual work in manufacturing. The new technology has promoted the introduction of a new
organisation of work with flatter organisational structures and job loss in middle management.
18
Conclusions and policy recommendations
J. Philpott also suggests that a possible answer to the puzzle of greater self-reported concerns
of increased job insecurity in the 1990s, despite lower unemployment levels, may be related
to the ongoing process of structural and occupational change. There are also signs that trends
towards greater wage inequality are, at least in part, driven by new technology premiums.
The main lesson from G. Schmid’s chapter, ‘Beyond conventional service economies —
utility services, service product chain and job services’, emanates from the decisive difference
between manufacturing and services. Whereas services provide immediate utilities to known
clients, manufactured products provide potential and standardised utilities sold on an
anonymous market. If service utilities are standardised and quickly consumed like fast food,
window cleaning or transporting beer, the valuing of these utilities can be also standardised
and is most efficiently organised by the market. The majority of services, however, are utilities
which are hard to measure. Their real value is uncertain, especially if they contain investment
elements like health, childcare, education, information and research services. The utilities of
such services often accumulate in the future, or they are realised only after some years or, in
the worst case, never. Most importantly, many service utilities are often realised only if they
are combined with other services or with non-monetary inputs from the consumer or client.
So, the problem of these kinds of services is how to organise a proper valuing system which
takes into account the value of the real outcome or performance which contains high risks and
uncertainty, and which also takes into account the often required non-monetary work to tap
the potential utility of high quality services. Risk management and risk sharing is the real
economic problem of the new service economy. The main thesis of G. Schmid’s essay,
therefore, is that those societies that display successful employment performance are those
which have established effective and efficient risk management systems.
Policy recommendations
The point of departure for policy recommendations is that this research confirms that job
creation in Europe will indeed be in the service sector. Our policy recommendations will be
made with reference to existing EU employment policy, and to the Employment Guidelines
and Structural Funds in particular.
The realisation that the increase of service employment in the EU is still accompanied by a
corresponding decline in agricultural and industrial employment, leads one to suggest policy
for job creation in the service sector in the broad context of structural change. The ongoing
shifts in employment structure, out of agriculture and manufacturing and into services, have
been accompanied by far ranging changes in the service sector. The new information and
communication technology could be seen as being the first major technological innovation
that has directly addressed large parts of the service sector. This has radically altered the skill
requirements of the labour force and has been a driving force behind the new organisation of
work. It has also increased the possibility to separate the production and consumption of
services, both as regards timing and geographical location.
19
The job creation potential of the service sector in Europe
In general terms, policy should serve to assist this ongoing transformation. Labour shed in one
sector or region of the economy is not automatically absorbed in expanding sectors or regions.
A logical policy prescription to facilitate labour market adjustment, is the promotion of labour
mobility, both occupational and geographical. Education and training policy has always been
used as a means of adapting labour to structural changes. This policy perspective is even more
relevant when one places the creation of jobs in the service sector with the rise of knowledgeintensive work. Hence, one of the most basic requirements for job creation in the new
economy is an increase in the supply of knowledge workers and flexible internal workplace
structures that can make the best use of knowledge workers. In the new economy, workers
benefit most from transferable skills that offer access to work, the ability to benefit from
lifelong and work-based learning, and to switch between a range of work functions and
careers. To achieve this, education and lifelong learning policies need to instil the ability to
use and impart knowledge. This implies the teaching of information management, personal
networking skills and problem-solving skills. In other words, education and initial training
should focus primarily on the provision of generic skills rather than narrow specific academic
or vocational qualifications. Several of the employment guidelines contained within the
improving employability pillar are consistent with this requirement.
The guidelines do not, however, explicitly set labour market policy as part of a model of
structural change. As agricultural (and some older traditional industrial) activities are
intrinsically regionally concentrated, geographical mobility should be a particularly important
priority. Legislation on freedom of movement has a long history and is a policy area that, for
good reasons, the European Union has a strong mandate. The Treaty of Rome dealt with the
right to mobility, the single market programme made some progress in enabling mobility
(qualifications and social security) and recent initiatives have attempted to deal with
information problems. Perhaps it is time to go a stage further with measures to actively
promote labour mobility. Some Member States currently provide both cash mobility grants
and moving cost assistance within their national borders.
The regional studies in Part Two, and in particular the United States experiences, underline the
importance of the regional and industrial aspects of structural policy. The research showed that
services cannot only generate net income and employment in their own region, through trade,
but also generate jobs in other sectors in the region. Moreover, the German experience shows
that, it is by no means exclusively business services that can play such a role. Social and
communal services (for example, education, health and cultural events) and distributive
services (Internet retail) can generate income from outside the region. The research also
showed that the growth in service employment is sensitive to some of the same location
factors as manufacturing and emphasises that service firms can choose their location. The
Structural Funds have practically no stipulations that the funds should award higher priority
to traditional industrial activities. Policy implementers at the local level often stress instead
the importance of attracting businesses that have a market even outside the region. Given that
services can increasingly constitute the economic base of the regional economy, this should
serve to remove any remaining regional policy bias against service as opposed to industrial
jobs.
20
Conclusions and policy recommendations
There is thus a case for industrial and regional policy to play a key role in the creation of jobs
in the service sector in Europe. Here it is important to underline the geographic, demographic
and structural diversity of the Member States. The major part of the EU shortfall of service
jobs is to be found in the southern, more agricultural countries, and should be a major focus
for policy. Moreover, the previously mentioned policy prescription of labour mobility can
hardly be a feasible, or indeed a desirable, means of fully shifting the geographical and
structural distribution of jobs and workers in Europe. At least part of the required regional
shifts will have to be made in the allocation of jobs and not workers. An important part of
European regional policy, directed to the relatively underdeveloped parts of Europe, has been
the financing of mainly transport infrastructure. As modern information and communication
technology permits a widening of the geographical distance between production and
consumption of services, there is a very strong argument for infrastructural investment in
modern communication and information networks in the more peripheral (and low service
intensive) regions. This would, of course, also require a complimentary investment in the
regional human capital.
It is also important not to underestimate the potential of creating service jobs that exploit the
comparative advantages of these regions — recreation, tourism, agri-environmental retail etc.
The major shortfall of service jobs in these regions is, however, in social and communal
services, and in health and education in particular. As the budget constraints that accompanied
EMU may limit the public financing of these services, there is reason to suppose that the
potential for service jobs organised around the social economy may be particularly appropriate
in such regions.
The realisation that services are more intrinsically tradable than previously thought, partly as
a result of the new technologies, should lead to renewed policy initiatives to further promote
their tradability within the Union. Economies of scale and the division of labour are a prime
motor for economic growth. To the extent that these jobs are driven by economies of scale, the
lack of a fully developed single market for services is an obvious candidate for explaining the
European shortfall in services. The policy prescriptions are largely along the lines of the 1985
White Paper on completing the internal market, namely, those concerning public procurement,
the right of establishment and other matters of company law and occupational qualifications.
Particular emphasis should be placed on facilitating marketing and internationalisation for
small and medium-sized businesses. Moreover, the creation of a large internal market for
services provides the basis for successful European services in the world market. It is only
through a large single market that competitive pressure and technical optimal scale may
develop and thus facilitate a globally competitive European service sector.
Improving the competitiveness of high quality and expert-oriented business services, and
relieving the cost burden for personal or labour intensive services are two recurrent policy
proposals for increasing employment in services. Whereas competitiveness of business
services can be increased through innovations in information and communication
technologies (innovative milieus or networks), it has been suggested that low skill personal
services be stimulated by widening the wage spread combined with various types of wage
subsidies or income policy.
21
The job creation potential of the service sector in Europe
Employment in business services can, as mentioned previously, be promoted through
education and training. However, this is not sufficient to raise productivity, growth and
employment without the adoption of flexible working arrangements and greater product and
capital market flexibility, which is why the guidelines on developing entrepreneurship and
encouraging adaptability are equally as important. At present entrepreneurs in the United
States find it easier to obtain venture capital than their EU counterparts and face a regulatory
environment which makes it relatively easy to set-up, and operate. It is thus appropriate that
the developing entrepreneurship guidelines encourage Member States to improve the
availability of venture capital, to reduce overhead costs and administrative burdens on
business, and to reduce tax obstacles to small business formation.
One other promising policy orientation is the promotion of innovative milieus or networks at
the regional level. However, as put forward by various authors of this report, the effective
governance of networks is still not well understood. They stress the following points; Firstly,
most effective regional networks are voluntary private enterprise networks in which policymakers have a very limited role to play. The most important support of these networks is
indirect through provision of ‘hardware’ infrastructure, such as excellent transport and
communication networks and ‘software’ infrastructure such as skilled and competent labour
force and information clearing. Secondly, some effective regional networks are public-private
partnerships in which regional governments or municipalities enter as co-financiers and,
therefore, as important bargaining partners. City contracts, regional pacts or target-specific
associations (in which local or regional governments are members) are common elements in
successful public-private partnerships. Thirdly, policy-makers can play an important role as
‘gatekeepers’ or moderators in the formation of regional networks. In playing the role of
‘moderators’, local or regional governments might take over an entrepreneurial role which
may indirectly lead to new jobs in new service markets. Fourthly, networks are no panacea.
Conventional strategies of economic policy, namely, setting proper market frameworks and
compensating obvious market failures and social policy (setting proper frameworks of social
security and equal opportunity policies), are still required.
Social networks play also a crucial role in adjusting to structural changes in the labour market,
especially in finding a new job. The policy consequences of this insight have not yet been
rigorously addressed. Although generous cash benefits can be regarded as financial
precondition to maintain networks (membership in clubs, going to social events with friends
etc.), a more proactive orientation towards establishing or supporting existing networks can
be taken. Some of this ‘experience knowledge’, of which the local placement officers are often
unaware, can be stimulated by arranging suitable context conditions, which usually involve
professional services related to job search. The activating of cash benefits into job (search)
services also makes sense from the knowledge that service jobs, especially in the new services
related to the information sector, are often ill-defined and changing fast. Thus, the importance
of experience knowledge is itself related to the service economy and its development.
Concrete policy proposals are the implementation of targeted vouchers or earning credits,
intensification of placement services, support of local community networks, buying
professional placement, training or rehabilitation services, and the support of new risk
22
Conclusions and policy recommendations
management services related to job creation — for instance business start-ups for the
unemployed.
As far as the promotion of labour-intensive personal services is concerned, the general
strategy is clear: a reduction of labour costs by means of wage subsidies, income policy or tax
differentiation. The extension of basic income support to persons with low earning capacities,
the professionalisation of household related services and the alleviating of cost for providing
such services are possible routes to stimulate both the demand and the supply of personal
services. However, the option to be selected should be left to the European Union’s Member
States, taking into consideration path dependency and slow learning capacities of institutional
routines. Part of this objective is being pursued by the employability guidelines on activation,
and by the developing entrepreneurship guidelines relating to the possibility of lower labour
taxes for low-wage labour and reductions in VAT on labour-intensive services.
A large part of the discrepancies in employment rate between the EU and the United States
can be ascribed to differences in female employment rates. The fact that women’s
commitment both in terms of employment rate and working time is, on average, higher in the
United States, means that the needs of household related services is greater than in European
countries where a large part of these activities are still performed within the household. In
other words, part of the EU–US employment rate gap can be ascribed to shortage and
availability of household related services, which constrain women’s labour supply. Hence, a
policy aiming at fostering an expansion of household related and social services may have a
beneficial effect on female labour supply (both in terms of decision to work and working
hours), gender equal opportunities and on female employment opportunities, as this kind of
service activity is still predominately performed by women. Part of this objective, that is, an
overall reduction of gender employment gap and the promotion of an adequate provision of
good quality care for children and dependants, is pursued by the equal opportunity guidelines.
The large variation in the relative size of the public and market services in Europe as well as
the large variation in women’s employment rates have also several policy implications. First,
this diversity means that there are differences in the potential for female employment growth
in traditional female employment areas, thereby leaving scope for public intervention. Hence,
public policies promoting the development of employment in social services (health,
education) and the ‘outsourcing’ of female domestic tasks into the market (childcare in
particular) may accelerate women’s labour market integration by transferring some tasks from
the domestic sphere to the labour market. This development would have a positive impact on
the overall employment rate and reduce the prevailing gap between the European Union and
the United States. In particular, the tight public expenditure restrictions of the EMU
constrains the scope for countries with underdeveloped welfare systems to ‘catch up’ with
other European countries. Consideration should be given to providing scope for public
expenditure expansion for service provision in these countries. More generally, consideration
should be given to ways of encouraging the restructuring of existing public expenditure away
from transfers to ‘breadwinners’ to support their ‘dependent’ spouses and into spending on
service provision, to improve the long-term employment and fiscal base.
23
The job creation potential of the service sector in Europe
The second policy implication is that any expansion of services that is heavily dependent on
female labour will draw more women into employment but do little to reduce the sexsegregated structure of employment. A two-pronged policy to address segregation is therefore
required. Firstly, anti-discrimination and affirmative action measures are required to enable
women to move into the better quality areas of male-dominated employment. Secondly, the
quality of the jobs which women currently do also needs to be improved. Positive action
should therefore not simply involve measures to move women into male-dominated areas, it
should also involve a reconsideration of the organisation of the job hierarchy to provide
greater recognition of the skills involved in many female-dominated job areas, and to create
links from these job areas into mainstream promotion lines.
The third issue is the under-representation of women among the self-employed. The share of
women engaged as family workers is declining as their involvement in waged work rises, but
there is less evidence of a trend towards equal representation in self-employment. On this
basis, action is needed to boost women’s involvement in entrepreneurship, particularly in the
southern countries where such activities are a major part of economic life. Enabling women
to set up businesses may mean that they develop businesses to deliver childcare and other
household related services. This may contribute to the expansion of such market services as
well as providing women with better opportunities for developing their own autonomous
careers.
‘In the end, the wealth producing forces of an economy are its immaterial resources — not oil
or gas, copper or iron ore, but rather the intellectual and organisational skills that are the very
core of the services’ (Krommenacker, 1984, p. 9).
24
Introduction
Dominique Anxo and Donald Storrie
The European Union has established an integrated framework for economic and employment
policies designed to improve its capacity to create jobs. Exploiting the opportunities for job
creation in the service sector appears as a major issue in the European employment strategy.
In the 2000 Employment Guidelines, the European Commission stressed the need to develop
a policy framework in order to fully exploit the employment potential of the service sector, in
particular by identifying and removing those obstacles which continue to hamper the growth
of enterprises and the creation of more and better jobs, and by using the employment potential
of the social economy and information society.
The issue of the job creation potential of the service sector in the European Union is
empirically illustrated with reference to the lower rate of employment in the service sector in
the EU compared with the United States. Despite a long period of increased service
employment share relative to the United States, full convergence has yet to be reached and the
EU share of total employment in services is still 8 % units lower than in the United States. In
1985, this gap was 12 %. The distribution of employment by economic sector is a
fundamental measure of the state of economic development, with development being
associated with the transition from the agrarian, to the industrial, to the service economy.
Despite clear evidence of European catch-up to United States level of service employment
share, the ranking of countries by service share is very much the same today as it was over at
least 40 years ago.
Due to the recent higher growth of the aggregate employment rate in the United States
compared to the EU, the recent growth of employment rate by economic sector gives a
somewhat different impression of the relative growth of services in the EU and the United
States. Between 1985 and 1997 the absolute increase in employment rate in services in the
25
The job creation potential of the service sector in Europe
United States was 7.8 %, while in the EU it increased by 4.7 %. However, while the total
difference in the employment rates in the United States and the EU was 5.8 %, 1.7 % of this
gap was attributable to the agricultural sector, 1.1 % to industry and 3.2 % to services. Thus,
the smaller agricultural and industrial sectors accounted for roughly the same gap change as
services. Recent EU and United States employment comparisons do not testify to a poor
European job creation in services compared to the United States. Poor EU job creation is an
across-the-board phenomenon.
However, even if one should temper impressions of poor EU service employment
performance implicit in the guidelines to this report, it still must be considered correct that
future employment growth in the EU economy will come from service jobs. The more
developed Member States have followed the stylised development from an agrarian to an
industrial and finally to a service-dominated employment structure. There would appear to be
no reason to suppose these countries will not continue to exhibit an expanding service sector.
The major part of the EU service employment gap, however, is not to be found in these
countries, but rather in the less developed Member States. However, it is highly unlikely that
these countries are following their more developed neighbours by first moving to a
predominantly industrial employment share. Indeed, the stylised development of first shifting
to industrial employment is unique to early western European development, and both the
United States and Japan (and other more recently developed economies) shed labour out of
agriculture mainly to the service sector. Thus, the basic premise of this report, that there is a
large job creation potential in the service sector is almost certainly correct.
The main objective of this research project is to understand the theoretical basis of the
dynamic of employment growth in services. The following report is structured around three
main themes. Part One, ‘The job creation potential of the service sector: General
considerations’, addresses the main empirical, theoretical, methodological and gender issues.
Part Two, ‘The industry and service nexus: Concepts and empirical evidence’, focuses on the
interaction between manufacturing industry and producer services in the dynamics of job
creation. Finally, Part Three, ‘New perspectives of the service economy’, reflects upon the
impact of the new information and communication technology for our conception of services
and the potential for job creation resulting from advances of this technology. Furthermore, the
job creation potential of services is examined in the light of the possible switch from product
market to utility markets and innovative possibilities in managing specific income risks
related to modern services.
In his contribution, ‘Service employment, productivity and growth’ (Chapter One, Part One),
D. Storrie collects empirical evidence in the United States and EU Member States regarding
the dynamic of structural changes and the job creation performance. First, it is necessary to
reflect upon the specificity of service activities. Early economists focused on the transient
nature of services which in the words of Adam Smith ‘perish in the very instant of their
performance.’ Consequently, some classical economists viewed services as essentially
unproductive labour. While this is hardly the modern view, it is just this perishable nature of
services that make the measurement of the value of services, and consequently productivity,
very difficult indeed. Later classical economists, most notably John Stuart Mill, however,
26
Introduction
could make the distinction between material and immaterial ‘perishability.’ He observed that
while some personal services, such as education and health, were materially perishable, this
did not mean that the consumption of a service lacked lasting utility and indeed could have
productivity enhancing effects in the production of goods. Moreover, the fact that services are
often materially perishable has the consequence that production and consumption occur
simultaneously and in the same geographical location. This chapter also introduces the
conceptual difficulties in analysing the output of service activities both as regards their
heterogeneity and intangibility. It is shown that these conceptual difficulties have concrete
empirical implications and result in very severe measurement problems associated with the
measurement of output and thus of productivity growth.
As far as the general conceptual level is concerned, the standard model of economic
development identifies two fundamental transitions, namely from agriculture to industry and
from industry to services. The terms primary, secondary and tertiary also indicate a
chronological progression. A large service sector is thus in some meaning viewed as being a
feature of an advanced or developed economy. The recurrent broad explanation of service
growth is based on two hypotheses. First, that income elasticity for the demand for services is
higher than for goods, i.e. that with increasing income an ever-larger share is devoted to the
consumption of services. Second, that there is a higher productivity growth in the production
of goods relative to services due to the more limited opportunity of embodying technical
progress in the latter. This is the basic assumption of Baumol’s cost disease model.
In Chapter 2 Part One, L. Tronti, R. Sestini and A. Toma, ‘Unbalanced growth and
employment in the services’, explore the reasons of employment expansion in the services,
stemming from the theoretical framework, represented by the Baumol model of ‘unbalanced
growth.’ The Baumol model is particularly interesting for studying the behaviour of
employment in the service sector, since service industries are generally characterised by slow
productivity growth, while agriculture and manufacturing show sustained long-term
productivity increases. The theoretical analysis of the Baumol cost disease mode proceeds by
successively relaxing the underlying assumptions of the model to explore the theoretical and
policy consequences.
One of the most salient features and persistent trends in advanced economies is the increased
feminisation of the labour force and the related shift from the male breadwinner household
towards dual earner households. The overall increase of female employment rates over the last
decades has been closely associated with the development of the welfare state in general and
the growth of services in particular. As roughly 80 % of all female employment in the
European Union is to be found in the service sector, it is obvious that any analysis of the
service sector must focus on the gender dimension. Despite the significant reduction of the
gender employment gap, there are still large differences in the patterns of female integration
in Europe. It is highly likely that the further integration of women in the labour market will
occur in the service sector. The focus of the contribution of D. Anxo and C. Fagan, ‘Service
employment: A gender perspective’ (Chapter 3, Part One) is threefold. Firstly, to analyse the
macro level relationship between the organisation of the welfare state, the employment
structure and the position of men and women in the labour market. Secondly, to explore the
27
The job creation potential of the service sector in Europe
implication of sex-segregated employment and service sector restructuring for women’s
labour market integration. Finally, to discuss the theoretical implication of policy designed to
shift or ‘outsource’ tasks which are largely carried out by domestic work and selfprovisioning, mostly by women’s labour, in the service sector of the economy. Such a
development could be expected to have implications for the gender division of labour and
female career opportunities, and for the level of employment growth and overall efficiency of
the economy.
While mainstream economists have focused their analysis on labour market rigidities (wage
rigidities and differentials) in explaining the employment deficit in the service sector between
the EU and the United States, the explanation of the employment gap may also be ascribed to
different social habits or norms and possibly reinforced by institutions and regulations. For
instance, institutional restrictions on the start up companies, capital market imperfections,
product market regulation or restrictive store opening hours may hinder job creation in many
countries. Identifying cross-country differences in the product market environment for the
service sector could be one route to assess national disparities in the dynamic of job creation
in the service sector.
The contribution of F. Benhamou and B. Gazier, ‘Labour costs, social habits and employment
regimes’, (Chapter 4, Part 1) addresses these issues by comparing the retail trade in France
and the United States and the cultural sector in France and the United Kingdom. Special focus
is placed on the societal, organisational and institutional context surrounding the work
relationships and the risk management arrangements in explaining the country relative
employment performance of these sectors.
The second part of the report, ‘The industry and service nexus: Concepts and empirical
evidence’, focuses on the job creation potential of services in relation to other sectors of the
economy, in particular with manufacturing. In empirical terms, producer services are highly
significant as the largest EU–US employment gap is to be found in this sector. It is also the
fastest growing of all services and is the major current source of European catch-up.
The dynamics of the service sector and the relative importance of the various factors in
explaining service sector growth may also benefit from an analysis performed at the regional
level. Regional analysis is particularly useful, as it is at this level that the relationship between
the manufacturing and service sectors can best be explored. For instance, regions in the
Member States or the United States are characterised by concentrations of various industries.
Therefore, exploring the relationship between manufacturing and services in various regions
allows the identification of the reasons why some service sectors are growing faster than
others. A further heuristic advantage with the regional approach is that it makes it possible to
identify whether service employment growth is limited to particular regions or sectors, or is
ubiquitous across all services and regions. Previous research suggests that service growth in
the EU and the United States is centred in high growth areas and in specific business sectors
such as communication and information services. The empirical basis of the first two chapters
of Part II rely on regional data to analyse the manufacturing/service nexus.
28
Introduction
The chapter of R. Exerts and G. Erickcek (Chapter 1, Part II), ‘The nature and determinants
of service sector growth in the United States’, concentrates on the following main areas: After
a brief look at employment trends and a short review of some of the relevant literature, the
authors explore the shift in composition of occupations within and between services and
manufacturing. This analysis provides insight into the changing nature of the two industries
in terms of the functions that workers perform and also the extent to which manufacturing has
outsourced service-related functions. One of the primary issues related to outsourcing is the
use of temporary help and other flexible workforce arrangements. The authors discuss the
relationship between the growth in business services and this change in working relations, and
examine the sources of demand for services as revealed through the interaction of
manufacturing and service firms, as well as the final demand of non-business consumers.
The contribution of D. Dathe and G. Schmid, ‘The determinants of business and personal
services, evidence from the German regions’, theoretically identifies the factors explaining
the contrasting employment growth in personal and producer services in Germany, and also
explores the relationship between employment in the manufacturing and the service sector.
Whereas mainstream economics and comparative sociology rely almost completely on
national regimes to explain differences in the performance of service employment, the authors
assume a relative autonomy of regional actors from framework conditions at the national
level. This is not to deny the determination power of national institutional frameworks, but the
regional approach exploits a neglected source of variation which gains increasing plausibility
considering the globalisation process. Using this regional variation, the authors test the
empirical validity of the manufacturing/service nexus.
Previous empirical evidence suggests that manufacturers are outsourcing more and more of
their ‘non-core’ functions. At the micro level the theoretical basis of the analysis of whether a
firm performs the service in-house or purchases it on the market is to be found in the theories
concerning the boundaries and structure of the firm. The principle focus has been on
transaction costs, economies of scale and scope and general problems of coordination in
situations of uncertainty. An issue here is the degree to which outsourcing will result in
employment growth in the service sector net of the job loss in the industrial sector as defined
in national accounts. One could speculate that the same type of work that becomes outsourced
will require less labour than before due to, for example, economies of scale. Rather
paradoxically perhaps, the job creation potential of business service outsourcing may be
expected to be in the industrial sector, as it presumably now becomes more competitive with
obvious job creation effects. The purpose of the chapter by B. Gazier and N. Thevenot,
‘Analysing business services employment — some theoretical and methodological remarks’,
is to explore some important dimensions and determinants of the growth of business services.
The authors begin by analysing business services from a purely managerial point of view
using the service profit chain approach. They stress, in particular, the importance of trust and
loyalty for business services, where the creation of long-term relationship between the
employees and the customers may be mirrored by a correspondingly long-term relationship
between the employees and the employer. Business services and outsourcing are further
analysed within the transaction cost framework and also by combining a labour, industrial and
organisation economics perspective. Finally, the chapter addresses some methodological
29
The job creation potential of the service sector in Europe
issues, in particular the measurement difficulties and the lack of reliable data for estimating
the outsourcing process.
The final part of this report, ‘New perspectives of the service economy’, takes up some more
innovative perspectives of the service economy. As indicated in the 2000 Employment
Guidelines, information technology plays a vital role in the growth of service employment.
The new information and communication technology has already had dramatic effects on
various service sectors. This technology could be seen as being the first major technological
innovation that has directly addressed parts of the service sector, as many services often entail
the processing of information. References to the ‘new economy’ are nowadays commonplace
and the underlying theme is that developed economies are increasingly knowledge dependent,
founded on information and communications technology, and ever more geared to the
production of services rather than primary and manufactured goods. In addition, technological
improvements have altered the physical attributes of manufactured goods, thereby making
them lighter (miniaturisation being the most obvious example). Moreover, far more of the
value of goods and services resides in the higher quality that richer and more sophisticated
consumers demand of them, notably in the form of design and specification. This has led to
the new concept of the ‘weightless economy.’ The labour market corollary is that the new
economy is only capable of creating weightless work. Fewer workers are employed in the
production of tangible goods and those that remain are increasingly knowledge workers —
needed to improve the intangible qualities of products — rather than either unskilled muscle
workers or routine skilled or unskilled manual operatives. This in turn leaves the service sector
— which produces intangibles — as the sole source of net employment growth. Chapter 1 in
Part III by J. Philpott, ‘Weightlessness and the political economy of services sector job
creation’, discusses the causes and consequences of weightless growth and assesses what this
means for employment and the labour market. The author then considers the related broad
policy implications and sets out criteria that policy-makers might wish to take into account
when deciding upon the relative merits of policies designed to create service jobs in the
private, public and social economy sectors. Two pathways to increase service employment
are usually suggested: improving competitiveness of high quality and expert-oriented
business services, and relieving the cost burden for (low skill) personal services. Whereas
competitiveness of business services can be increased through innovations in information and
communication technologies (innovative milieus or networks), it is suggested that low skill
personal services be stimulated by widening the wage spread combined with various forms of
wage subsidies or income policy. There are still untapped possibilities to improve these
strategies by policy reforms or further research. However, both pathways are rather
conventional and do not exhaust the range of innovative approaches towards the modern
service society.
In the last chapter of Part III, ‘Beyond conventional service economics — utility services,
service product chains and job services’, G. Schmid explores three complementary ways to
foster job creation in services: First, the switch from product markets to utility markets would
induce more and new types of user services. Such a switch would not only be more
employment friendly but also foster a sustainable economic development compatible with
ecological principles. The reorientation from products to utilities provides strong arguments
30
Introduction
for an emphasis on the framework conditions of services, for instance for lowering VAT in
favour of labour intensive services, for forward regulation through (usually service
enhancing) high quality standards, for deregulating market entry barriers, and for subsidising
networks instead of individual products or services. Second, Baumol’s cost disease has to be
reconsidered in view of productivity enhancing possibilities by new technologies, and in view
of innovative possibilities in managing specific income risks related to modern services.
Service product chains are the main ingredient for increasing productivity in services, and
demonstrate this in the field of art services, Baumol’s prototype and the inspiring area for the
theory of cost disease. The evolution of social insurance in this field offers also inspiring ideas
for risk management beyond the traditional welfare state.
31
Part I
The job creation potential of the service sector:
General considerations
Chapter 1
Service employment, productivity
and growth
Donald Storrie
1.1. Introduction
The issue of the job creation potential of the service sector in the European Union has been
empirically illustrated with reference to the United States. Despite a long period of increased
service employment share relative to the United States, full convergence has yet to be reached
and the EU share of total employment in services is still 8 % units lower than in the United
States. In 1985, this gap was 12 %. The distribution of employment by economic sector is a
fundamental measure of the state of economic development, with development being
associated with the transition from the agrarian, to the industrial, to the service economy.
These are very long-term trends, and despite considerable convergence of service
employment share among the advanced economies, the ranking of countries by service share
is much the same today as it was over at least forty years ago.
Due to the recent higher growth of the aggregate employment rate in the United States, the
recent growth of employment rate by economic sector gives a somewhat different impression
of the growth of services in the EU relative to the United States. Between 1985 and 1997 the
absolute increase in employment rate in services in the United States was 7.8 % and 4.9 % in
the EU. However, while the total difference in the employment rates in the United States and
the EU was 5.8 %, 1.8 % of this gap was attributable to agricultural sector, 1.1 % to industry
and 2.9 % to services. Thus, the smaller agricultural and industrial sectors accounted for
roughly the same widening of the gap as did services. Recent EU and United States
employment comparisons do not testify to a particularly poor European job creation in
services compared to the United States. Poor EU job creation is an across-the-board
phenomenon.
35
The job creation potential of the service sector in Europe
However, even if one should temper impressions of poor EU service employment
performance, implicit in the guidelines to this report, it still must be considered correct that
future employment growth in the EU economy will be in services. The more developed
Member States have followed the stylised development from an agrarian to an industrial and
finally to a service dominated employment structure. There would appear to be no reason to
suppose that these countries will not continue to exhibit an expanding service sector. The
major part of the EU service employment gap, however, is found in the less developed
Member States. It is highly unlikely that these countries are following their more developed
neighbours by first moving to a predominantly industrial employment share. Indeed, the
stylised development of first shifting to industrial employment is unique to early western
European development, and both the United States and Japan (and other more recently
developed economies) shed labour out of agriculture mainly to the service sector. Thus, the
basic premise of this report, that there is a large job creation potential in the service sector is
almost certainly correct.
Section 1.2 presents a brief discussion of the concept of services. Services are a very
heterogeneous set of economic activities. Compared with the production of goods, the product
of services is difficult to define. While earlier definitions focused on this intangibility, more
modern definitions focus on the importance of the interaction between the provider and the
user. The matter of intangibility is, however, of great practical importance and is the key to
understanding the severe problems encountered in the measurement of service output and
productivity discussed in section 1.4.2.
Section 1.3 describes the long-run development of the growth of services as outlined above.
For more recent years, and for a more detailed breakdown of service employment, data from
the European labour force survey (ELFS) is used. Throughout this section reference is made
to the US–EU gap in service employment. The subcategories of the service sector used in this
chapter are producer, personal, distributive and social. The largest current EU–US service
employment gap is found in producer services (and in all its subcategories), and personal
services (mostly attributable to a larger hotels and restaurants sector in the United States). A
smaller gap can be found in distributive services. The gap that does exist is entirely due to
higher United States employment in trade and repair. While on average, social and communal
services are roughly the same in the United States and the EU, the European figures are held
up by a higher level of employment in public administration. The United States has, however,
a higher share of employment in health and social services than the EU average. Recent trends
in service employment by service subsector show an appreciably higher growth of producer
services in Europe both in terms of employment share and rate.
The EU has of course by no means a unified employment structure. The personal services and
trade and retail gap is much lower in the less developed economies of Europe. The producer
service gap is much lower in the more developed countries. There is most EU variation in the
level of social services, in particular in health services.
Section 1.4 outlines the very long-term development of the employment structure in
agriculture, industry and services. As these shifts are often seen as being driven by
36
Service employment, productivity and growth
productivity differentials, the discussion of service productivity is also taken up here. It is
shown that the measurement problems, particularly of service output, are so severe that one
has reason to question the empirical validity of economy-wide productivity statistics. This
problem, serious as it is, is becoming continually more acute as service employment continues
to constitute an ever increasing share of total employment. The assumption of low service
productivity, which is the cornerstone of the Baumol model, cannot be taken as being
empirically valid. The final sub-section concludes with a brief examination of the new growth
literature for evidence of the growth enhancing effects of services. Research and development
activities, education and health are the obvious candidates for growth enhancing services.
Some research also finds positive effects for producer services.
1.2. The concept of services
While in everyday conversation one refers to services without too much reflection on their
meaning, other than as a function performed by someone for someone else, a precise and
useful definition is surprisingly difficult. While one should not over-emphasise conceptual
issues in this context, some reflection on the matter can serve to highlight some of the
theoretical and practical difficulties when analysing services.
An examination of the different services reveals a very heterogeneous range of economic
activities. Moreover, one finds that services are an exceedingly elusive concept. The main
source of difficulty is to identify what exactly it is that is being produced. This basic
conceptual issue leads to serious empirical problems in measuring output and productivity. As
service jobs account for between 60 and 70 % of total employment in modern advanced
economies, this has serious consequences for the quality of modern national accounts
(Griliches, 1994).
The perishable nature of services, economic activities that in the words of Adam Smith ‘perish
in the very instant of their performance,’ was that which occupied the thoughts of early
economists. While perishibility or intangibility is a characteristic of some services, the
tendency of classical economists to thus view service jobs as unproductive labour is not the
modern view and is hardly correct. This early theoretical disdain, fuelled perhaps by the
identification of the economic growth of recent centuries to a narrow definition of ‘industry,’
still in some quarters sustains a view of services as somehow of less importance. For example,
much more effort is still devoted to the collection of statistics on agriculture and industry than
on services (1). Later economists, most notably John Stuart Mill, could distinguish between
material and immaterial perishability. He observed that while some services are materially
perishable, like education and health, this did not mean that they lacked lasting utility. Indeed,
modern growth theory ascribes great importance to the benefits of intangibles, such as
education and health (and maybe even finance).
(1) Girliches (1992) points out that in the United States, the compilation of health and education statistics receives
less than half of the funds spent on agricultural statistics. One would expect that the situation is hardly better in
the European Union.
37
The job creation potential of the service sector in Europe
However, many services are undoubtedly in some sense perishable or intangible. This
intangibility means that the definition and measurement of service output is very difficult. It
also means that unlike most goods they can seldom be stored. Neither can they be easily
transported and thus are less traded. The necessity for a direct interaction between the provider
(producer) and the user (consumer) has consequences for the structure of the market and may
lead to spatial monopoly.
While some services lack tangibility or material substance a definition of services in terms of
material perishability raises numerous exceptions. Even a haircut makes material changes and
lasts for a while. A better definition focuses on the interaction of the provider and the user.
While even the distinctiveness of this definition is eroding with the new information and
communication technology (2), it still is the most accepted definition of services. It was
defined, as in Hill (1977), ‘A service may be defined as a change in the condition of a person,
or of a good belonging to a person belonging to some economic unit, which is brought about
as the result of the activity of some other economic unit, with the prior agreement of the former
person or economic unit.’
In terms of classifying services as an economic sector in, for example, the national accounts,
definitions may vary slightly from country to country but in most cases services fall within the
economic sectors not classified as agriculture or industry.
1.3. Employment in services
1.3.1. Introduction
This section traces the growth of the service sector share of employment for as many countries
and for as long a time period as the data will allow. This is an essential measure for describing
the shifting structure of employment with economic growth. A number of sources are used.
Long run aggregate data is based on the organisation for economic cooperation and
development (OECD) compilations. The long-term growth of a disaggregation of services is
presented using previous research (Elfring, 1988) and the European labour force survey
(ELFS) is used for a detailed description of the development during the 1990s.
As this report is to address the job creation potential of the service sector in light of the
superior United States development, the presentation of data for the EU countries will often
be expressed in terms of the gap with the United States. The first sections express the growth
of services as a share of total employment. The process of the shift from agriculture and
industry is a useful way to study the process of job creation in the service sector. In the latter
sections we present data on employment in the sector as a percentage of the working age
population, that is, the employment rate. This latter measure jointly captures the total job
creation of the economy and the structural shift.
(2) For example, Internet retail and possibly virtual reality advice services.
38
Service employment, productivity and growth
1.3.2. The long-run growth of aggregate service employment share
In 1960, the United States of America, together with Canada and Australia, had by far the
largest share of employment in the service sector. In that year the United States of America
figure was 56 % while the EU average was 39 %, see Table 1.1. Considerable variation was
found within Europe, with Denmark, the Netherlands, Sweden and the United Kingdom all
over 44 %. Greece, Spain and Portugal lay all around, or below, 30 %. We note that in 1960,
the service share in developed Germany was rather low and that it was high in the relatively
underdeveloped Ireland, both 39 %.
Since 1960 there has been a continual convergence towards the high United States levels. The
convergence rate has been higher in the EU than in OECD in general. Table 1.2 presents the
same figures as in Table 1.1 but focuses on the narrowing of the share gap between the United
States and the EU countries, Japan and Canada. The countries are sorted by size of service
share in 1960.
We observe the greatest absolute convergence in the four countries that had the lowest service
share in 1960, that is, Greece, Spain, Portugal and Finland. However, considerable
convergence is also found for France and Sweden. A striking feature of the data in Table 1.2
is that the highly successful Japanese and German economies, which were in the middle of the
ranking in 1960, have converged very slowly towards the United States service share. The
pattern of convergence is thus rather mixed. On the one hand, there is an across-the-board
convergence towards the high United States service sector share. On the other hand, by the
end of the period, it was still Canada, Belgium, Denmark, the Netherlands, Sweden and the
United Kingdom which had the highest service share, that is, the same six as in 1960. It
appears that, at least to a certain extent, the reasons for the current service share gap are to be
found in features of these economies 40 years ago.
39
The job creation potential of the service sector in Europe
Table 1.1. Service employment share in OECD countries 1960–95
Belgium
Denmark
Germany
Greece
Spain
France
Ireland
Italy
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
Canada
Japan
US
EU-15
OECD (EUR)
OECD
1960
1968
1974
1985
1990
1995
46.4
44.8
39.1
25.5
31.0
39.9
39.0
33.5
49.7
37.1
24.8
32.2
44.0
47.6
54.1
41.3
56.2
39.0
36.9
43.1
51.2
49.9
43.0
32.6
36.5
45.8
41.7
39.3
54.1
43.1
34.1
41.1
49.8
51.3
59.7
45.7
59.4
44.3
42.0
48.0
55.2
58.0
46.3
36.2
39.6
49.9
44.6
43.2
58.4
43.4
31.3
47.6
56.4
55.1
63.1
50.1
63.4
47.9
45.5
52.1
66.7
65.2
54.4
43.7
49.9
60.4
55.2
55.2
67.0
52.9
42.2
56.5
65.3
63.0
69.5
56.4
68.8
57.4
54.4
60.1
68.9
66.9
56.7
48.4
54.8
64.6
56.4
58.8
69.1
55.2
47.6
60.6
67.3
66.0
71.1
58.7
70.9
60.7
57.6
62.8
71.4
68.1
59.3
56.4
60.8
68.9
60.7
60.3
73.7
60.0
56.4
64.6
71.0
70.5
73.0
60.7
73.1
64.3
60.6
64.6
Source: OECD (1996).
Table 1.2. Service employment share gap with the United States of America 1960–95
Belgium
Denmark
Germany
Greece
Spain
France
Ireland
Italy
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
Canada
Japan
United States
Source: OECD (1996).
40
1960
1968
1974
1985
1990
1995
Narrowing
of the gap
9.8
11.4
17.1
30.7
25.2
16.3
17.2
22.7
6.5
19.1
31.4
24.0
12.2
8.6
2.1
14.9
56.2
8.2
9.5
16.4
26.8
22.9
13.6
17.7
20.1
5.3
16.3
25.3
18.3
9.6
8.1
– 0.3
13.7
59.4
8.2
5.4
17.1
27.2
23.8
13.5
18.8
20.2
5.0
20.0
32.1
15.8
7.0
8.3
0.3
13.3
63.4
2.1
3.6
14.4
25.1
18.9
8.4
13.6
13.6
1.8
45.9
26.6
12.3
3.5
5.8
– 0.7
12.4
68.8
2.0
4.0
14.2
22.5
16.1
6.3
14.5
12.1
1.8
15.7
23.3
10.3
3.6
4.9
– 0.2
12.2
70.9
1.7
5.0
14.0
16.7
12.3
4.2
12.4
12.8
– 0.6
13.1
16.7
8.5
2.1
2.6
0.1
12.4
73.1
8.1
6.4
3.1
14.0
12.9
12.1
4.8
9.9
7.1
6.0
14.7
15.5
10.1
6.0
2.0
2.5
16,9
Service employment, productivity and growth
We conclude this section with some remarks on the quality of this long-run aggregate data.
Generally speaking, adherence to the International Labour Organisation (ILO) definitions is
better for the more advanced economies and improves over time. The different treatment of
multiple job holders, the self-employed and unpaid family workers are the main problems.
Moreover, one has reason to believe that these problems will be most severe for services (due
to the large number of small and family firms in services). As the broad aggregate services is
generally classified as not being in primary or secondary sectors, the sector definition problem
(adherence to ISIC) should not be too serious.
A more serious problem is that of equating service work with the service sector. For example,
according to Tschetter (1994), in 1992, one third of all workers employed in United States
manufacturing were actually doing service type jobs, for example, in finance, purchasing,
marketing and administration. Moreover 14 % of the workers in United States service
industries had production type jobs, for example, construction workers and mechanics.
However, the relatively stable distribution of occupations by economic sector appears to
exclude such examples as statistical illusion to explain the increase in service employment
(United States Bureau of Commerce).
1.3.3. The long-term growth of service employment share by subsector
As was pointed out earlier, services are a heterogeneous set of economic activities and for most
analysis one requires a more detailed aggregation. Due, above all, to the different classification of
employment by economic sector, there exists no acceptably consistent set of statistics published
by international bodies. This is particularly the case as regards older data. Thus, at this level of
aggregation, the presentation of the long-term trends must rely on previous research. The sector
classification (distributive, personal, producer and social) is that proposed by Hill and utilised by,
for example, Singelmann (1978), Elfring (1988) and Castells (1996). The period 1960 to 1984 is
well covered by Elfring (1988), who standardised definitions using national sources. Elfring
presents statistics for the United States of America, Germany, France, the Netherlands, Sweden
and the United Kingdom. In some respects this is an interesting choice as they can be seen as an
appropriate sample of the world’s most advanced economies. The choice was also based on the
availability of statistics of sufficient quality. It would probably not have been feasible to present
statistics of sufficient quality before the advent of the European labour force survey, for example,
for Greece, Spain, Ireland, and Portugal.
For the European countries, the figures from 1995 onwards are based on the European labour
force survey. For the United States we use data from the current population survey. The sector
distribution (producer, personal, distributive and social services) in this data is as close as we
could attain in accordance with the definitions used by Elfring (1988). We cannot document
the degree of compatibility of our recent figures with Elfring’s standardisation as regards the
treatment of other issues. However, a glance at the data shows no obvious inconsistencies.
Table 1.3 shows the level of the service sector share of total employment for all services and
by the four major subcategories for six countries from 1960 to 1997 (3). In 1960, the United
(3) The definition of these subsectors is apparent from Table 4.
41
The job creation potential of the service sector in Europe
States of America had the largest service sector share in each of the four categories, with the
largest absolute gap in social services. However, by 1997 the US–Europe gap had been
reversed with none of these European countries having a lower social service share of total
employment than the United States. The case of Sweden is the most remarkable, with a very
large increase between 1973 and 1984. For other European countries there has been a more
gradual but continuous catch-up and subsequent take-over during the whole time period.
Table 1.3. Service employment share by subsector in six countries
All services
Year
1960
1973
1984
1997
Germany
France
Netherlands
Sweden
United Kingdom
United States
38.6
44.2
47.8
47.6
48.8
61.1
46.1
51.3
57.7
57.7
55.3
66.4
53.1
60.6
67.3
66.4
64.9
71.8
62.0
68.5
73.1
70.9
70.9
73.6
Distributive
Year
Personal
1960
1973
1984
1997
1960
1973
1984
1997
Germany
France
Netherlands
Sweden
United Kingdom
18
17
20
19
21
18
19
21
20
20
18
20
22
19
21
20
20
23
19
22
7.4
7.9
8.5
8.4
8
6.5
7.5
7.6
6.6
7.9
7.4
7.7
7.8
5.9
9.7
7.1
8.2
6.8
5.7
9.2
United States
22
22
21
22
11
11
12
12
Producer
Year
Germany
France
Netherlands
Sweden
United Kingdom
United States
Social
1960
1973
1984
1997
1960
1973
1984
1997
3.4
3.5
4.2
3.5
4.4
6.4
5.2
6
6.8
5.1
6.5
8.7
6.7
7.9
9.9
6.4
9.2
12
11
12
14
12
14
15
10
16
15
16
16
21
16
19
23
16
21
25
21
25
28
35
25
26
25
29
30
34
25
25
Source: For 1997, EU countries — ELFS, 1997, United States — from Employment in Europe, 1998.
For earlier years Elfring, 1988.
The other large service sector, distribution, has also exhibited convergence towards the United
States of America employment share and by 1997 lay between 19 and 22 % of all employment
in all countries. However, the small relative differences in the distribution services
employment share are important in absolute terms and warrant further examination later in the
next section.
42
Service employment, productivity and growth
In 1960, the number employed in personal services was greater than in producer services. Up
to 1973 the personal service share of all countries declined. This is possibly due to a
substitution of personal for social services. The fact that Swedish personal service share
declined even up to 1997, a period during which social service share more than doubled, is an
indication of such a substitution. For the other countries, there has been a small increase since
1973. The level of social services has remained remarkably stable over the entire period with
little convergence towards the stable United States level.
In several respects, producer services are the most interesting of all the service sectors. This
is the sector with the highest growth rates for all countries and despite being a smaller sector
than both the social and distributive sectors accounts for most of the US–Europe gap.
However, the growth of this sector in the European countries between 1984 and 1997 has been
very rapid (see in particular the Netherlands, Sweden and the United Kingdom) and an
extrapolation of the recent trend would soon eradicate this gap.
Since the mid-1990s, the general industrial classification of economic activities within the
European Communities (NACE) classification of sectors was fully implemented in the ELFS.
This enables a detailed classification of employment by economic sector of high quality for
all current Member States from 1995 onwards. The structure of employment in 1997 is
presented in Table 1.4 together with United States data based on the current population survey.
The 1998 data, for the EU only, are presented in Table 1.5.
Table 1.4 shows that the total EU service sector share is 8 percentage units lower than the
United States with both a larger agricultural and industrial share. Employment structure within
the EU varies greatly. Structure in the Netherlands is almost identical to that of the United
States. In terms of service share, the European countries sort into two groups. Belgium,
Denmark, France, Luxembourg, the Netherlands, Sweden and the United Kingdom have less
than around a 5 % gap to the United States. Much larger gaps are found in Germany, Greece,
Spain, Ireland, Italy, Austria, Portugal and Finland. A major part of the gap for this latter group
is accounted for by an appreciably higher agricultural sector. The major exception is Germany,
which has the largest industrial share of all countries.
Nearly all of the aggregate EU gap to the United States is attributable to the lower average
European share in producer and personal services. For the low service sector share of
European countries, the gap is largely accountable to the low level of social and producer
services. The three lowest service intensive nations in the EU have share of personal service
very similar to the United States. On the other hand, the more service intense European
countries, have an appreciably higher share of social services. The producer services gap is
much lower and the personal services gap is much higher.
43
44
D
EL
E
F
L
NL
IRL
I
A
P
FIN
S
UK
– 1.21
2.32
5.05
3.51
0.81
0.73
4.58
– 0.08
0.73
1.64
0.40
1.90
– 5.93
– 5.32
– 1.30
0.66
0.03
– 0.96
0.14 – 1.32
3.22
2.30 – 0.05
5.42
4.55
2.01
3.96
3.59
0.93
0.24
0.97
0.86
1.22 – 0.01
0.23
3.79
4.61
8.47
0.35
0.23
0.97
1.02
0.90
1.24
1.11
1.37
1.99
– 0.10
0.56
1.05
1.41
1.55
3.21
– 7.01
0.06
6.77
– 1.71 – 4.43 – 2.79
0.18
2.38
1.73
– 5.24
2.20
7.07
– 0.24 – 0.09
0.75
– 0.73
– 0.03
– 0.12
0.71
0.58
– 1.41
6.30
0.89
1.16
1.68
0.84
1.73
6.29
– 2.03
1.78
5.92
0.63
– 0.34
3.14
3.41
3.61
0.76
– 0.96
3.37
0.43
1.17
0.75
0.38
0.66
– 4.00
– 4.81
0.19
0.99
– 0.37
– 1.32
2.93
2.74
1.72
1.24
– 0.22
– 1.37
– 6.58
1.01
1.78
0.84
1.57
– 3.23
– 5.21
1.28
4.22
– 3.51
– 0.52
0.48
0.05 – 0.75
0.94 – 1.70 – 0.59
0.09
2.51 – 0.09
0.90
2.28
5.00
4.27
4.83
1.36
3.04
2.63
0.83
5.11
5.88
3.64
1.33
2.49
1.20
2.06
4.03
4.24
0.36
0.11
1.65
0.94
1.18 – 0.05
0.12
0.84 – 0.08 – 1.10
0.50 – 2.42
1.13
1.52
1.06
5.30
6.22
4.62
7.51
4.46
2.86
0.15
0.13
0.35
0.19
0.76
1.04
1.14
1.05
0.93
0.97
0.64
1.25
1.02
1.31
1.32
1.68
1.66
1.03
1.74
0.76
0.32
0.04
0.39
0.38
0.79
0.86
0.19 – 0.23
– 1.50
2.17
2.85
1.98
2.91
1.43
0.30
– 4.95
3.32
3.13
3.21
6.13 – 3.39 – 9.30
– 3.50 – 0.79 – 3.17 – 2.23 – 2.18 – 0.79 – 1.02
1.18
1.05
0.10
1.80
0.84
0.72
0.37
– 2.82
2.90
5.73
3.66
6.91 – 3.09 – 8.15
0.19
0.16
0.48 – 0.01
0.57 – 0.23 – 0.51
Source: EU — author’s compilation of ELFS; US — the current population survey, from Employment in Europe, 1998,
– 0.28
1.59
3.24
2.80
0.68
– 0.24
4.06
0.14
1.10
1.21
0.41
1.21
– 0.07
– 3.19
0.96
2.07
0.09
– 0.44
1.05
2.45
2.26
– 0.30
0.49
0.80
– 1.15
1.44
0.69
0.07
– 0.26
– 0.61
– 1.51
0.22
0.54
0.13
3.92
16.62
11.62
6.89
2.43
2.30
15.14
2.84
1.89
2.03
1.22
7.16
24.73
4.46
7.70
11.49
1.08
DK
— Transport
— Trade and repair
Personal services
— Hotels and restaurants
— Recreation and culture
— Other personal
Producer services
— Banking
— Insurance
— Real estate
— Computer
— Other producer
Social services
— Public administration
— Education
— Health and social
— Other social
B
2.57 – 2.40 – 0.09 – 1.17 – 0.37 – 17.27 – 5.73 – 2.08
0.23 – 1.14 – 8.34 – 3.96 – 4.32 – 10.72 – 5.21 – 0.68
0.71
23.92 – 5.88 – 3.90 – 2.51 – 11.14
1.08 – 6.24 – 2.92
0.29
0.75 – 4.81 – 8.28 – 6.12 – 7.52 – 3.82 – 1.89 – 3.36
73.85
8.42
4.12
3.81 11.65 16.33 12.11
5.13 – 0.39
0.52 13.29 12.38 10.58 18.38
9.16
2.70
2.79
22.16
1.19
0.42
1.61
2.43 – 0.91 – 0.36
2.34
1.47 – 0.42
3.31 – 0.01
0.11
3.90
2.99
3.26
0.16
1.62 – 0.12 – 0.69 – 0.65 – 0.01
0.46
0.41 – 0.46 – 0.13
0.01
0.32
0.02 – 0.04
0.68 – 0.31 – 0.42 – 0.45
EU
European gap from the US level
Agriculture
Industry
Services
Distributive services
— Communications
US
Level
Table 1.4. EU employment share gap to the United States, 1997
The job creation potential of the service sector in Europe
4.77
29.88
0.38
20.59
7.72
1.19
65.35
20.94
1.71
4.24
14.99
8.40
4.05
1.77
2.57
11.28
2.62
0.81
0.84
0.92
6.09
24.74
7.46
6.81
9.46
1.01
Source: Author’s compilation of ELFS data.
Agriculture
Industry
— Mining
— Manufacturing
— Construction
— Electricity, gas, water
Services
Distributive services
— Communication
— Transport
— Trade and repair
Personal services
— Hotels and restaurants
— Recreation and culture
— Other personal
Producer services
— Finance
— Insurance
— Real estate
— Computer
— Other producer
Social services
— Public administration
— Education
— Health and social
— Other social
EU
2.24
27.49
0.26
19.41
6.60
1.21
70.28
21.58
2.24
4.72
14.62
6.75
3.46
1.69
1.60
11.62
3.13
1.30
0.41
1.02
5.76
30.32
9.55
8.89
10.80
1.09
B
3.71
26.67
0.15
18.84
6.74
0.95
69.62
21.08
2.29
4.77
14.03
5.79
2.76
2.10
0.94
11.38
2.37
0.66
1.02
1.58
5.76
31.36
5.95
7.13
17.08
1.20
DK
2.78
34.80
0.53
24.12
8.83
1.32
62.42
19.74
1.54
3.86
14.34
7.02
3.16
1.48
2.39
10.80
2.55
1.01
0.74
0.74
5.76
24.85
8.55
5.54
9.54
1.23
D
17.75
23.41
0.46
14.57
7.12
1.27
58.84
23.00
1.18
4.99
16.83
10.30
6.28
1.68
2.34
7.32
1.81
0.62
0.06
0.19
4.64
18.21
7.02
6.07
4.67
0.45
EL
7.91
30.75
0.43
19.50
9.85
0.97
61.34
22.43
1.21
4.65
16.57
11.75
6.05
1.82
3.88
9.00
1.79
0.71
0.37
0.53
5.61
18.16
6.25
6.02
5.44
0.45
E
4.42
26.61
0.21
18.88
6.43
1.09
68.96
19.89
2.07
4.35
13.47
8.26
3.23
1.73
3.29
11.90
2.38
0.79
1.26
0.88
6.58
28.92
9.33
7.73
10.48
1.37
F
2.92
21.97
0.16
12.40
8.51
0.90
75.11
19.70
1.72
5.08
12.90
8.39
4.44
1.05
2.89
17.85
9.51
1.14
0.37
0.46
6.38
29.17
10.25
6.14
7.28
5.51
L
3.49
23.15
0.18
15.81
6.12
1.03
73.36
23.00
1.58
4.70
16.71
6.45
3.40
1.79
1.27
14.97
2.62
1.06
0.80
1.46
9.02
28.95
7.20
6.72
13.99
1.04
NL
10.91
28.73
0.47
19.13
8.06
1.07
60.36
18.86
1.31
3.44
14.11
10.27
5.56
2.32
2.38
9.84
2.71
0.96
0.35
0.82
4.99
21.40
5.25
6.65
8.58
0.92
IRL
Table 1.5. Employment share by economic sector, 1998
6.44
32.11
0.47
22.62
7.62
1.40
61.45
21.98
1.56
3.76
16.66
8.59
4.32
0.78
3.48
9.07
2.46
0.89
0.38
0.85
4.49
21.81
7.76
7.49
5.97
0.60
I
6.48
29.94
0.30
20.46
7.78
1.41
63.58
22.34
1.65
4.74
15.95
9.15
5.72
1.62
1.82
10.48
2.62
1.26
0.95
0.67
4.98
21.61
6.49
5.99
8.03
1.10
A
13.74
36.43
0.34
24.27
10.77
1.05
49.83
17.55
0.93
2.75
13.86
10.61
5.29
1.02
4.30
5.44
1.34
0.47
0.25
0.30
3.09
16.24
5.81
5.84
4.08
0.51
P
7.11
28.49
0.23
20.66
6.08
1.52
64.40
18.72
2.86
5.45
10.41
6.13
2.77
2.23
1.12
11.26
1.48
0.71
1.34
1.20
6.53
28.30
5.28
7.23
14.05
1.74
FIN
3.06
26.19
0.24
19.49
5.31
1.16
70.75
19.39
2.18
4.40
12.81
5.86
2.67
2.47
0.72
12.12
1.69
0.50
1.54
1.56
6.83
33.38
5.30
7.41
19.19
1.48
S
1.73
27.11
0.37
18.59
7.07
1.08
71.17
21.85
1.96
4.58
15.31
9.25
4.58
2.89
1.78
14.73
3.92
0.49
1.33
1.37
7.62
25.34
5.83
7.61
10.96
0.94
UK
Service employment, productivity and growth
45
The job creation potential of the service sector in Europe
From Table 1.4 we see that the EU-15–US gap in distributive services is 1.2 percentage units.
Given the absolute size of this sector this corresponds to a rather small relative gap. Moreover,
a breakdown to the subsectors shows that it is only in the large trade and repair sector that
employment is greater in the United States. The trade and repair gap is, however, small for the
five least service intensive EU Member States.
The aggregate EU-15–US gap in social and communal services is negligible. Every EU nation
State has more employees in public administration. This should hardly be surprising as we
compare one national administration with 15 national administrations. The United States has
a larger share of its employed persons in health and education than the EU average. However,
considerable variation can be found within the European Union. The Nordic countries and the
Netherlands employ more in the health and social service sector than the United States. A more
unified picture is apparent as regards employment in education with the United States
employing more in education than 14 of the EU countries.
The United States employs more in personal services than all EU countries with the exception
of Spain. The average gap is three percentage units and given the relatively small size of the
sector this should be considered an appreciable gap. A breakdown into subsectors reveals that
the gap is due to larger United States employment in hotels and restaurants. This gap is much
less apparent in the least service intensive EU countries
The largest absolute gap is found in producer services. The gap is quite evenly distributed
among the various subsectors with no single subsector accounting for the EU-15–US gap. It
is, however, notable that the relatively small real estate sector exhibits quite a large gap. This
gap is appreciably higher in the less-service-intensive EU countries. Indeed, the variation of
producer services among these nations is closely correlated to GDP per capita and is the only
service subsector to exhibit such a correlation.
1.3.4. Recent growth of service employment rates
In this section we focus on the growth of employment rates in services, that is, the number
employed in a sector as a percentage of the working age population.
The ELFS data presented in Tables 1.4 and 1.5 above should be deemed as being of very high
quality as it is based on a common source. This data is, however, only available for the EU from
1995 to 1998. Nonetheless, the 1998 employment rates report presents data between 1985 and
1997. Using this data we calculate the change in employment rates by service subsector.
Unfortunately, the service classification is not as useful as the one used by Elfring and others.
Due to all of the above matter of sector aggregation, this data is probably not of the highest
quality. For example, the very rapid increase in the Greek figures appears somewhat strange.
However, it is of interest to present the broad development both in terms of employment
structure, as in previous sections, and in rates, which is the focus of this section.
Table 1.6 shows the recent shifts in aggregate employment structure and rates in the EU
countries compared to the United States between 1985 and 1997. Looking first at structure,
46
Service employment, productivity and growth
we see that the EU, compared to the United States, was still shedding agricultural labour and
in relative terms this is the major difference between the two. The industrial structure gap
remained the same during this period, with EU industrial sector share six percentage units
higher than the United States. Consequently, the EU has increased its share in services relative
to the United States (this is the employment share convergence trend shown for a longer
period in Table 1.2). Thus, in terms of the European gross flows, relative to the United States,
labour lost in agriculture moved to services.
The rates data in the top part of the table give quite a different impression as regards the
service gap. In 1985 the absolute gap was 12 percentage units and by 1997 it had widened to
15 %. The difference between the two figures, the rates and the share is of course attributable
to the greater increase in the aggregate employment population ratio in the United States.
Table 1.6. Employment share and rates by broad economic sector in the United States
and EU-15, 1985 and 1997
EU
United States
Gap
Absolute
Relative (%)
Employment rate
Agriculture
Industry
Services
Total
1985
1997
1985
1997
1985
1997
1985
1997
5
3
21
18
34
39
60
61
2
2
19
18
46
54
68
74
–3
–1
–2
–1
12
15
8
14
– 58
– 35
– 8.0
– 3.0
35
38
13
22
Employment structure
Agriculture
Industry
Services
1985
1997
1985
1997
1985
1997
8
5
34
30
57
65
3
3
28
24
69
73
–5
–2
–6
–6
12
8
– 63
– 40
– 18
– 20
21
12
Source: Compilation of data from the employment rates report, 1998.
Table 1.7 presents the growth of employment rates between 1985 and 1997. We see that the
absolute increase in the employment rate of services was 7.8 % in the United States and 4.9 %
in the EU. However, in Greece, the Netherlands, Austria and Portugal (4), service employment
rates exhibited a larger increase than the United States.
(4) As noted earlier, I suspect that the data are not of the highest quality and am particularly sceptical about the
Greek data.
47
The job creation potential of the service sector in Europe
Table 1.7. Absolute change in employment rates by broad economic sector
and service subsector, 1985–97
Agriculture Industry
Services Distribution Transport
Producer Communal
Belgium
Denmark
Germany
Greece
Spain
France
Ireland
Italy
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
EU-15
United States
– 0.4
– 2.1
– 0.8
– 6.7
– 4.0
– 2.0
– 2.3
– 2.5
– 0.2
– 1.3
– 4.8
– 4.1
– 1.9
– 0.3
– 1.9
– 0.1
– 0.9
– 0.7
– 3.1
– 4.2
0.7
– 3.5
1.4
– 1.7
– 1.1
– 5.1
– 0.5
– 5.8
– 6.3
– 4.0
– 2.3
– 1.2
5.4
4.0
2.5
10.3
7.8
3.4
6.8
2.3
10.4
8.7
9.8
– 0.3
– 3.4
8.8
4.7
7.8
0.6
0.9
– 0.8
8.2
2.1
– 0.2
2.0
– 0.6
3.8
2.8
3.4
– 1.5
– 0.5
1.0
0.6
1.1
0.3
0.1
0.0
– 0.8
0.3
– 0.3
0.4
0.0
0.2
– 0.1
0.1
– 0.5
– 1.3
0.6
0.1
0.4
1.8
3.1
1.2
1.4
1.6
1.2
1.2
2.6
3.6
3.6
3.3
1.6
2.7
4.0
2.2
1.5
2.6
– 0.1
2.2
1.6
3.7
2.7
3.3
0.2
2.8
2.1
3.1
0.1
– 4.4
3.2
2.0
4.7
Japan
– 2.1
0.6
5.8
0.8
0.5
1.7
2.8
Source: Compilation of data from the employment rates report, 1998.
We also note that, in the important sector of producer services, even when expressed in terms
of employment rate, that growth is higher in the EU. Job losses in European agriculture and
industry were much larger than in the United States. Indeed, one could claim that the European
performance in relation to the United States was worse in these sectors than in services.
Table 1.8 presents the rates data in 1998 by broad economic sector and according to the
producer, distributive, personal and social service classification. Table 1.9 presents the change
in employment rates by disaggregated economic sector according to the same sector
classification. The change figures are expressed as both an absolute change and relative
change between 1995 and 1998. The most striking feature of this table is that producer
services have, by far, the highest growth rates. This is also the case in nearly every Member
State.
48
DK
D
EL
E
F
L
NL
IRL
I
A
P
FIN
S
UK
1.90
2.91
1.29
2.94
1.71 10.15
3.97
2.69
1.77
2.38
6.60
3.33
4.55
9.46
4.63
2.15
1.23
17.70 18.26 15.82 21.13 21.41 13.39 15.44 16.18 13.29 15.80 17.37 16.62 20.99 25.09 18.54 18.41 19.34
54.50 39.94 40.44 55.16 38.40 33.66 30.80 41.94 45.44 50.08 36.51 31.81 44.57 34.31 41.91 49.73 50.79
26.00 38.89 42.45 20.77 38.48 42.80 49.79 39.19 39.50 31.74 39.52 48.24 29.89 31.14 34.92 29.71 28.64
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
16.40 12.80 12.42 16.71 12.14 13.16 11.26 12.10 11.92 15.70 11.40 11.38 15.66 12.08 12.18 13.63 15.59
1.20
1.04
1.29
1.81
0.95
0.67
0.61
1.26
1.04
1.08
0.79
0.81
1.16
0.64
1.86
1.53
1.40
2.90
2.59
2.72
3.78
2.37
2.85
2.33
2.64
3.07
3.21
2.08
1.95
3.32
1.90
3.54
3.09
3.27
12.30
9.16
8.41 11.11
8.82
9.63
8.32
8.19
7.81 11.41
8.53
8.62 11.18
9.54
6.77
9.01 10.92
8.60
5.13
3.88
4.59
4.32
5.89
5.90
5.02
5.07
4.40
6.21
4.44
6.42
7.30
3.99
4.12
6.60
5.10
2.48
1.99
2.18
1.95
3.59
3.04
1.97
2.69
2.32
3.36
2.24
4.01
3.64
1.80
1.88
3.27
1.80
1.08
0.97
1.66
0.91
0.96
0.92
1.05
0.64
1.22
1.40
0.40
1.13
0.70
1.45
1.73
2.06
1.70
1.57
0.92
0.75
1.47
1.34
1.95
2.00
1.75
0.86
1.44
1.80
1.27
2.96
0.73
0.50
1.27
11.20
6.89
6.69
9.02
6.65
4.19
4.52
7.23 10.80 10.21
5.95
4.70
7.35
3.75
7.33
8.52 10.51
2.10
1.60
1.80
1.88
1.57
1.04
0.90
1.45
5.75
1.79
1.64
1.28
1.84
0.92
0.97
1.19
2.80
1.40
0.50
0.75
0.52
0.62
0.35
0.36
0.48
0.69
0.73
0.58
0.46
0.88
0.32
0.46
0.35
0.35
1.50
0.51
0.23
0.81
0.45
0.04
0.18
0.77
0.22
0.54
0.21
0.20
0.67
0.17
0.87
1.08
0.95
0.90
0.56
0.59
1.25
0.45
0.11
0.26
0.53
0.28
1.00
0.50
0.44
0.47
0.21
0.78
1.10
0.97
5.30
3.72
3.32
4.57
3.54
2.66
2.82
4.00
3.86
6.16
3.02
2.32
3.49
2.12
4.25
4.80
5.44
18.30 15.12 17.45 24.85 15.29 10.42
9.12 17.58 17.65 19.76 12.95 11.29 15.15 11.18 18.42 23.46 18.08
3.30
4.56
5.49
4.72
5.26
4.02
3.14
5.68
6.20
4.91
3.17
4.02
4.55
4.00
3.44
3.72
4.16
5.70
4.16
5.11
5.65
3.41
3.47
3.02
4.70
3.71
4.59
4.02
3.88
4.20
4.02
4.70
5.21
5.43
8.50
5.78
6.21 13.53
5.87
2.67
2.73
6.38
4.40
9.55
5.19
3.09
5.63
2.81
9.14 13.49
7.82
0.80
0.61
0.63
0.95
0.76
0.26
0.22
0.83
3.33
0.71
0.56
0.31
0.77
0.35
1.13
1.04
0.67
B
Source: EU — author’s compilation of ELFS; United States — the current population survey (from Employment in Europe, 1998).
Agriculture
Industry
Services
Not working
Total
Distributive services
— Communication
— Transport
— Trade and repair
Personal services
— Hotels and restaurants
— Recreation and culture
— Other personal
Producer services
— Finance
— Insurance
— Real estate
— Computer
— Other producer
Social services
— Public administration
— Education
— Health and social
— Other social
US(1997) EU
Table 1.8. Employment rates in the EU by economic sector, 1998
Service employment, productivity and growth
49
50
DK
D
EL
E
F
L
1.24 0.03 1.46 0.04 3.94 0.08 – 0.08 0.00 2.09 0.07
0.03 0.00 – 0.42 – 0.03 0.60 0.04 – 0.70 – 0.05 0.28 0.02
– 0.03 – 0.03 0.04 0.03 0.20 0.12 – 0.16 – 0.14 0.06 0.09
0.03 0.01 – 0.25 – 0.09 – 0.06 – 0.02 – 0.18 – 0.07 – 0.19 – 0.06
0.03 0.00 – 0.21 – 0.02 0.47 0.04 – 0.37 – 0.04 0.42 0.05
0.24 0.05 0.28 0.08 0.38 0.09 0.09 0.02 0.61 0.12
0.10 0.04 0.13 0.07 0.34 0.19 0.07 0.04 0.29 0.09
0.10 0.10 0.14 0.16 0.18 0.12 0.14 0.18 0.05 0.06
0.04 0.02 0.01 0.01 – 0.14 – 0.16 – 0.12 – 0.07 0.27 0.25
0.57 0.09 1.04 0.18 1.47 0.19 0.34 0.05 0.63 0.18
– 0.03 – 0.02 0.20 0.12 0.11 0.06 – 0.12 – 0.07 0.06 0.06
0.01 0.02 0.05 0.08 0.06 0.14 – 0.06 – 0.08 – 0.03 – 0.07
0.03 0.07 – 0.04 – 0.16 0.19 0.31 0.07 0.18 0.01 0.30
0.15 0.36 0.28 0.90 0.42 0.51 0.11 0.31 0.04 0.67
0.41 0.12 0.55 0.20 0.68 0.17 0.34 0.11 0.55 0.26
0.40 0.03 0.56 0.03 1.49 0.06 0.19 0.01 0.56 0.06
– 0.09 – 0.02 – 0.25 – 0.04 0.19 0.04 – 0.34 – 0.06 0.00 0.00
0.15 0.04 0.21 0.04 0.31 0.06 0.17 0.05 0.23 0.07
0.33 0.06 0.53 0.09 1.02 0.08 0.41 0.07 0.26 0.11
0.01 0.02 0.06 0.11 – 0.02 – 0.02 – 0.05 – 0.07 0.07 0.23
2.89 0.10
– 0.84 0.08
0.00 0.00
0.17 0.08
0.66 0.09
0.19 0.03
0.06 0.02
0.11 0.14
0.02 0.01
0.83 0.23
0.01 0.01
0.03 0.10
0.02 0.15
0.09 0.53
0.67 0.31
1.03 0.13
0.18 0.06
0.42 0.16
0.39 0.17
0.04 0.24
NL
0.86 0.02 3.03 0.07 3.80 0.08
0.06 0.00 – 0.66 – 0.05 1.05 0.07
– 0.04 – 0.03 0.11 0.12 0.03 0.03
0.11 0.04 – 0.07 – 0.02 0.37 0.13
– 0.01 0.00 – 0.70 – 0.08 0.65 0.06
0.19 0.04 – 0.23 – 0.04 0.14 0.03
– 0.05 – 0.03 – 0.31 – 0.10 0.01 0.01
0.07 0.07 – 0.11 – 0.15 0.02 0.02
0.18 0.10 0.19 0.12 0.10 0.14
0.20 0.03 2.04 0.23 1.75 0.21
– 0.01 – 0.01 0.68 0.13 0.22 0.14
– 0.02 – 0.04 0.11 0.18 0.19 0.35
0.04 0.05 0.02 0.10 0.06 0.13
0.02 0.04 0.15 1.21 0.44 0.78
0.18 0.05 1.08 0.39 0.84 0.16
0.41 0.02 1.87 0.12 0.86 0.05
– 0.01 0.00 0.34 0.06 – 0.36 – 0.07
0.14 0.03 0.50 0.16 0.43 0.10
0.24 0.04 0.42 0.11 0.66 0.07
0.03 0.04 0.60 0.22 0.13 0.23
Source: Author’s tabulations of ELFS. Note that there are no ELFS data for Ireland in 1998.
All services
Distributive services
— Communications
— Transport
— Trade and repair
Personal services
— Hotels and restaurants
— Recreation and culture
— Other personal
Producer services
— Banking
— Insurance
— Real estate
— Computer
— Other producer
Social services
— Public administration
— Education
— Health and social
— Other social
B
IRL
I
A
P
FIN
S
UK
3.17 0.10
1.16 0.11
0.08 0.11
0.28 0.15
0.81 0.10
0.63 0.11
0.33 0.11
0.11 0.08
0.20 0.16
0.35 0.06
0.06 0.04
0.03 0.06
0.05 0.32
0.02 0.04
0.19 0.07
1.03 0.09
0.07 0.02
0.24 0.06
0.72 0.16
– 0.01 – 0.02
1.10 0.04 1.89 0.04 – 2.34 – 0.06 1.77 0.04 – 1.17 – 0.02
0.06 0.00 – 0.04 0.00 – 0.48 – 0.04 0.48 0.04 – 0.09 – 0.01
– 0.01 – 0.02 – 0.09 – 0.08 – 0.14 – 0.18 0.40 0.27 – 0.07 – 0.04
0.03 0.02 – 0.09 – 0.03 – 0.19 – 0.09 0.40 0.13 – 0.09 – 0.03
0.04 0.00 0.14 0.01 – 0.16 – 0.02 – 0.32 – 0.05 0.07 0.01
0.22 0.05 0.46 0.08 0.86 0.13 0.45 0.13 – 0.04 – 0.01
0.08 0.04 0.40 0.11 0.62 0.20 0.24 0.16 0.11 0.06
0.04 0.10 0.15 0.15 0.12 0.21 0.20 0.16 – 0.13 – 0.07
0.10 0.06 – 0.09 – 0.06 0.13 0.04 0.01 0.02 – 0.02 – 0.04
0.54 0.13 0.88 0.14 – 1.07 – 0.22 0.83 0.13 0.31 0.04
– 0.01 – 0.01 – 0.06 – 0.03 – 0.52 – 0.36 – 0.25 – 0.21 0.06 0.05
0.02 0.03 0.14 0.18 – 0.12 – 0.28 0.16 0.51 – 0.02 – 0.07
– 0.03 – 0.13 – 0.02 – 0.03 0.02 0.13 0.09 0.12 – 0.12 – 0.10
0.10 0.31 0.16 0.50 – 0.02 – 0.09 0.19 0.31 0.38 0.52
0.46 0.25 0.67 0.24 – 0.43 – 0.17 0.64 0.18 0.01 0.00
0.28 0.03 0.59 0.04 – 1.65 – 0.13 0.02 0.00 – 1.35 – 0.05
0.06 0.02 – 0.03 – 0.01 – 0.85 – 0.17 – 0.26 – 0.07 – 0.2 – 0.01
– 0.03 – 0.01 0.10 0.03 – 0.73 – 0.15 0.14 0.03 – 0.07 – 0.01
0.21 0.07 0.43 0.08 – 0.09 – 0.03 0.13 0.01 – 1.22 – 0.08
0.04 0.13 0.08 0.12 0.02 0.07 0.01 0.01 – 0.04 – 0.03
2.24 0.05
0.23 0.02
0.05 0.03
0.18 0.06
0.01 0.00
0.34 0.05
0.15 0.05
0.16 0.09
0.02 –0.02
0.91 0.10
0.00 0.00
0.04 0.14
0.08 0.10
0.32 0.48
0.47 0.09
0.75 0.04
0.00 0.00
0.25 0.05
0.50 0.07
0.00 – 0.01
AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC AC RC
EU
Table 1.9. Absolute and relative change in employment rates by economic sector, 1995–98
The job creation potential of the service sector in Europe
Service employment, productivity and growth
1.3.5. Concluding remarks on the EU–US service gap by share and rate
Shifts in the distribution of employment by economic sector are the basic facts used to
describe the long-term development of the economy. Since 1985 the most striking relative
change in the sector distribution of employment in the United States and Europe is to be found
in the agricultural sector. In the United States, structural shift out of agriculture appears to
have fully run its course and agricultural employment has remained constant at 3 % between
1985 and 1997. In Europe, or rather in the less developed parts of Europe, this is still an
ongoing process and the agricultural share of total employment fell during this period from 8
to 5 %. While in the EU as a whole, agricultural employment amounts to only under 5 % of
all employment, it is still a very important source of employment in several Member States.
These are: Greece (18 %), Spain (8 %), Ireland (11 %), Italy (6 %), Portugal (14 %) and
Finland (7 %). Despite its size, the reduction of agricultural employment in Europe accounts
for 2 percentage units of the change employment rate differential in Europe and the United
States between 1985 and 1997. Presumably, the shedding of labour out of agriculture will
continue in the less developed European countries. The pace of this job loss will depend upon
the course of the common agricultural policy. Between 1985 and 1997, the shift out of industry
has proceeded at roughly the same rate in both the EU and the United States. Thus, in terms
of shares, there has been a narrowing of the service share gap to the United States in services,
from 12 percentage units in 1985 to 8 percentage units in 1997.
A key statistic behind the rationale of this report is that between 1985 and 1997, the
employment rate gap between the United States and the EU widened by 5.8 percentage units.
The sector components of the gaps were 1.9 % in agriculture, 1.1 % in industry and 2.9 % in
services. The gap growth in services is thus roughly of the same size as the sum of the other
two sectors. Thus, in relative terms, it is primarily the greater losses in the primary and
secondary sectors that have contributed to the poorer EU performance.
Despite convergence in European service employment structure towards the United States, a
gap of 8.5 percentage units still remains. In terms of the four major service sectors, the gaps
to the United States are:
Producer services
Personal services
Distributive services
Social services
4.1 %
3.2 %
1.2 %
0.1 %
At the finer level of aggregation the largest aggregate EU–US gaps are in the following
sectors:
Trade and repair
Hotels and restaurants
Real estate
Other producer services
Health and social services
1.6 %
2.8 %
1.2 %
1.2 %
2.1 %
51
The job creation potential of the service sector in Europe
Nearly all of the gap in personal services is due to the higher United States share in hotels and
restaurants. It is not at all obvious why this should be a matter of major EU policy concern in
terms of the overall productive capacity of the economy. While one should not make sweeping
statements about job quality in such a large sector, it is here that one finds the vilified
hamburger flipping ‘Mac-jobs.’ Indeed the European countries with the smallest gap in this
sector are the relatively underdeveloped Greece and Spain. Similar comments could be made
about the gap in distributive services. This gap is more than accounted for by the trade and
repair sector. The countries in Europe with the largest trade and repair sector are Greece, Spain
and Italy.
The European gap in other services is perhaps of more policy concern. While social services
constitute roughly the same employment share in Europe and the United States, it is the high
European employment in public administration that holds up the European figures. There is a
large European shortfall in health and social services. These are particularly low in Greece,
Spain, Ireland and Portugal. Moreover, even if the gap is small, it is notable that employment
share in education is lower than the United States in every European country, with the
exception of Belgium. As the new growth theories ascribe great importance to health and
education, the European shortfall should be seen as being worth further study.
The largest gap is found in producer services. While some may argue that this may be only a
statistical illusion, I do not believe this to be a pertinent response. R. Eberts and G. Erickcek
(Chapter 1 of Part 2 in this volume) show that while there is some evidence of a simple shift
of the same jobs from manufacturing to producer services, this by no means explains all the
empirical evidence and that net jobs are created. Moreover, even if it was the case that the
same jobs are now outsourced this outsourcing is presumably a more efficient division of
labour. The European gap in producer services is over 3 percentage units in all Member States
bar Luxembourg, the Netherlands and the United Kingdom and is apparent in all producer
subcategories. Producer service share of the employment varies directly with per capita
income (the only service subcategory with such a strong correlation). While one can certainly
find ‘bad’ jobs in this sector (cleaning and simple clerical work), it contains many high
productivity jobs associated with the rise of the information and communication technology.
Between 1985 and 1997 this was the fastest growing sector. It was shown in Table 1.3 that the
producer services gap between the United States and EU countries was previously rather
stable. Since 1985, however, there is a considerable European catch-up. This fact holds both
for employment structure (share) and rates. Within the EU, the ELFS data shows that between
1995 and 1998, producer services was by far the fastest growing sector. Presumably this trend
will continue and the EU–US gap will continue to diminish.
1.4. Services and economic development
1.4.1. Long-term shifts in employment structure
Changes in the economic structure as outlined in the previous section are the basic facts used
to describe the development from the agrarian, to the industrial and then to the service
52
Service employment, productivity and growth
economy. The terms primary, secondary and tertiary indicate a chronological progression and
have been viewed as synonymous with broad economic development and with income per
capita in particular.
‘Studying economic progress in relation to the economic structure of different countries we
find a very firmly established generalisation that a higher average level of real income per
head is always associated with a high proportion of the working population engaged in tertiary
industries. ... When we examine the trend over time (within countries) we find a similar result.
In every case we find the proportion engaged in primary industry declining and tertiary
increasing’ (Clark, 1940 pp. 6-7). This is also clearly demonstrated in Fuchs (1980).
While in all of the advanced economies, services is now the largest sector in terms of
employment, the path in terms of size relative to agriculture and industry has been quite
different. Table 1.5 presents data on the long-term development of employment shares. The
table shows that it is only the western European countries that have followed the stylised
movement from agriculture as the largest source of employment (pre-1870 for the United
Kingdom), to industry and finally to services as the largest employer. All countries have,
however, experienced the major change during the last 130 years, namely a dramatic fall in
agricultural employment. This has been extensively documented by economic historians, and
there is little doubt that this was due to a massive increase in agricultural productivity and the
low income elasticity of agricultural products. (Kuznets, 1966 and Chernery, 1979.)
The other common trend is the monotonic increase of the service sector. One can note that the
two European countries with the highest current service share, the Netherlands and the United
Kingdom, held the same ranking in 1870. All countries see first an increase and then a
decrease in industrial employment. However, in some countries, like the United States and
Japan, industrial employment was never larger than service employment, while in others, like
Germany and the United Kingdom, there was a period when the economy was an industrial
economy in terms of employment share. Industrial employment (as a share of the total) peaked
in the early 1960s in the United States of America (34 %). In the mid-1960s came the peak for
the Netherlands (40 %) and Sweden (38 %). Japan (37 %), Germany (50 %), France
(39 %) and Italy (39 %) peaked in the early 1970s. After these peaks there was a steady decline.
Table 1.10. Employment share in advanced economies share by broad economic sector
1870
Germany
France
Netherlands
Sweden
United Kingdom
Japan
United States
1960
1984
A
I
S
A
I
S
A
I
S
50
49
37
54
23
73
50
29
28
29
n.a.
42
n.a.
24
22
23
34
n.a.
35
n.a.
26
14
21
11
15
5
33
8
48
36
41
42
46
30
31
38
43
48
43
49
37
61
5
8
5
5
3
9
3
42
32
28
29
32
34
25
53
60
67
66
65
57
72
Source: For 1870, Maddison (1980); others from OECD (1996).
53
The job creation potential of the service sector in Europe
As early developers, the countries presented in Table 1.10 have been the object of most
previous research. Greece, Spain, Ireland and Portugal are much later developers. As seen in
Table 1.11 their agricultural share in 1960 was of roughly the same size as the more developed
countries in 1870.
Table 1.11. Employment share in the less advanced EU economies by broad economic sector
1960
Greece
Spain
Ireland
Portugal
1974
1985
1995
A
I
S
A
I
S
A
I
S
A
I
S
57.1
38.7
37.3
43.9
17.4
30.3
23.7
31.3
25.5
31.0
39.0
24.8
36.0
23.2
22.8
34.9
27.8
37.2
32.6
33.8
36.2
39.6
44.6
31.3
28.9
18.3
15.9
23.9
27.4
31.7
28.9
33.9
43.7
49.9
55.2
42.2
20.5
9.2
11.7
11.3
23.2
30.1
27.7
32.3
56.4
60.8
60.7
56.4
Source: OECD (1996).
Industrialisation came much later in these countries. The 1960s were a period of rapid
industrialisation and industrial share increased up to 1985, only then starting to decline. For
the early developers in Table 1.10, industrial share decline occurred much earlier. The later
developed countries never attained the same level of industrial employment share attained by
the early developers. Here, the structural transformation of employment has occurred even
more markedly from agriculture to services.
The future development of these countries is of great importance in this context as they
account for an appreciable part of the current gap between the EU and the United States. It is
highly unlikely that these countries will follow the pattern of Germany or the United
Kingdom. The United States development of the shift from agriculture to services is the most
probable development. It should be noted, however, that the very positive recent Irish
development has seen some increase in industrial employment. Between 1995 and 1998
industrial share rose in Ireland from 28 to 28.7 %, the corresponding change of service share
was from 60 to 60.4 %. While the differences are hardly sizeable, both the absolute and
relative increase has been greater in industry. The importance of industrial employment also
appears in the Greek structure. In Greece, industrialisation comes later than in the other
countries and is perhaps the least developed of the EU Member States. Industrial share has
declined since 1984 and has never reached 30 % of all employment (as it did in all other
countries).
1.4.2. Productivity in services
Productivity in services is a very important empirical, theoretical and policy issue.
• The fact that measured productivity in services has been relatively low, together with the
secular increase in service employment, is one of the major candidates for explaining the
slowdown of economic growth since the 1960s.
54
Service employment, productivity and growth
• Perception of low productivity growth in services is the crucial hypothesis in the dominant
theoretical explanation (The Baumol model) of the rise in service employment.
• If the EU is to promote the expansion of the service sector it is important to ascertain the
productivity of such jobs.
1.4.3.1. The measurement of productivity in services
Despite the importance of this issue, the serious problems in the measurement of, above all,
service output have yet to be satisfactorily resolved. Prominent economists have expressed
grave doubts as to the state of our empirical knowledge of service productivity. Paul Romer,
perhaps the principle architect of modern growth theory, claims ‘We do not know what
happened with the productivity slowdown in two senses. First, I don’t believe we know for
sure what the basic facts are. The quality of the data is such that we cannot speak with
authority and answer the question about what happened over time to the rate of growth of
productivity. Second, even if there was a slowdown we do not know the reasons with any
confidence’ (Snowdon and Vane, 1999).
Zvi Griliches is probably the world’s leading expert on the measurement of productivity. He
chose measurement problems in the service sector as the topic of his 1994 presidential address
to the American Economic Association. He concluded that by 1990 ‘… the fraction of the
economy for which productivity numbers are half reasonable had fallen to below one third...
Our ability to interpret changes in aggregate total factor productivity has declined and major
portions of actual technical change have eluded our measurement entirely.’ In other words, the
problem is bad and it is getting worse. The reason for Griliches’ perception of the increase in
the state of our ignorance is due to the continuing transformation of our economies out of
agriculture and industry and into services, where measurement problems are particularly
acute. Scepticism on the quality of recent United States productivity figures is also shared by
important United States economic policy-makers. In a recent speech on technology, Alan
Greenspan described much of the data on productivity as ‘non-credible’ and Treasury
Secretary Larry Summers thinks that statistical unreliability explains the lack of the measured
impact of the new technologies on productivity (5).
The source of the service measurement problems has already been mentioned and is due to the
material intangibility of service output as outlined in Section 1.1.2. When measuring the
output of goods over time, the usual procedure in the national accounts is to first ascertain the
money value of the output (value of shipments). This entails no conceptual or empirical
problems for services (revenues of services). The next stage in determining output for goods
is to divide this money value with a price index of the industry’s products. This requires a
specification of what exactly is being bought. For goods this is termed the basic transaction
unit and is a material entity. As there is seldom a clear-cut transaction unit for services, this is
the start of the empirical problems in measuring service output. The next stage in the
measurement of output for goods is to devise a means of measuring changes in the quality of
(5) Financial Times, 13 December 1999.
55
The job creation potential of the service sector in Europe
output. This is done by specifying the characteristics that determine the price of the good. If
this is done satisfactorily then one can distinguish price movements due to changes in quality
and quantity (output) may be adjusted accordingly (6).
What then is the unit of output for services? National account statisticians appear to embark
from Hill’s previously quoted definition of services as a change in the condition of a person,
or of a good, belonging to some economic entity, with the approval of the first person or
economic entity. As the result of the activity ‘a change in a condition’ there is often no tangible
unit that may be observed and counted. The degree of tangibility varies, however, between
different types of service. Advice services (legal, financial or medical) could be viewed as
fully intangible. Indeed the financial sector is very difficult to set in the framework of not only
outputs but also inputs where, for example, the treatment of customer deposits is very
problematic. In economic terms the output of educational services may be measured by the
number of students who pass exams. This is rather a crude measure of efficiency within
universities. The economic result of education will depend, among other things, on the type
of education. It will yield a stream of productivity for a long time into the future and will
almost certainly generate positive externalities. For these services the measurement of quality
changes will be even more difficult. In national accounts, many of these services are measured
in terms of the inputs. Obviously this gives very unsatisfactory measures of productivity.
Other measurement problems are associated with the availability of services. The word
‘service’ suggests the importance of the frequency or availability of services for its quality. For
consumers of, for example, health care, the availability of treatment should be an important
characteristic of its quality, given that the demand for health care is not fully predictable. The
eradication of queues will generally entail some overcapacity at non-peak periods, which
without a measure of availability will appear in statistics as low productivity.
One should, however, not generalise on the basis of these examples. Some service activities
generate more tangible outputs. For example, in transportation services, the material
definitional clarity of the service gives a relatively clear and quite easily measured output, that
is, the number of freight tonnes or number of passengers by the distance travelled and possibly
with some adjustment for the frequency or the time of the service.
When making international comparisons of productivity, further problems arise. While
comparisons of the levels of labour productivity between countries are of course of great
interest, one very seldom finds such data publicly published by the usual international
authorities (for example, organisation for economic cooperation and development (OECD),
international labour organisation, (ILO), and the United Nations, (UN)). Presumably the
reason for this is that the data is not of sufficient quality and is only published in terms of
productivity changes. When comparing aggregate output between countries, purchasing
(6) For example, it is quite feasible to capture the quality of the automobile in terms of, for example, horsepower,
petrol consumption, size and other concrete material measures. Of course there are measurement problems even
for goods. In the automobile example, it will probably be impossible to measure the ‘feel’ of different models.
Another often observed practical problem is that of time lags in the collection and compilation of this data by
statistical authorities in times of rapid technological change.
56
Service employment, productivity and growth
power parities (PPP) are used in an attempt to account for price differences. As prices vary
considerably from sector to sector, it is hardly appropriate to use the aggregate PPPs to deflate,
in this case, the service sectors. Moreover, casual observation leads one to believe that the
international price variation is greater for services than for goods. This is partly due to the
relative tradability of goods. Moreover, given that large sections of the service sector are not
priced at all, namely, public services, the problem is compounded further. Again the problem
varies from sector to sector and in the next section, comparisons for some sectors that more
easily lend themselves to international comparisons are presented.
It should also be noted that there are problems with measures of labour productivity (for
example, output per hour) for any subsector of the economy. Productivity improvements in
one sector may have considerable productivity benefits on other sectors. For example, the
major part of the productivity enhancing effects of technological progress in the transport
sector may spill over to the industrial or agricultural sector. Health, education and finance are
other obvious examples.
In general, there are few conceptual problems entailed in the measurement of inputs to
services (an exception is the banking industry where the role of customer deposits is
problematic). However, in practice there are greater problems in services than for goods, as
the statistical coverage of service establishments is considerably less than the coverage in
other sectors of the economy (7). This may be attributed to the historical lack of priority
awarded to the service sector mentioned earlier. Moreover, the relatively high incidence of
self-employed, part-time and unpaid family workers and the large number of small firms
makes data collection more difficult and costly.
1.4.3.2. Empirical evidence on service productivity levels and convergence
With the secular rise of service sector employment, the overall level of productivity in modern
economies is to an ever larger extent determined by productivity in the service sectors. This
fact, coupled with the measured low productivity of services is often viewed to lie behind the
economy-wide productivity slowdown since the 1960s.
With regard to this claim one should first set the slowdown in a long-term perspective. Two of
the major empirical studies of productivity growth in the last hundred years are Kendrick
(1973) and Maddison (1982 and 1987). Kendrick’s data on total factor productivity growth
show a highly pro-cyclical pattern, but no long-term trend. There is an indication of lower
levels since the 1960s but this is not at all remarkable in the long-term perspective. Similarly,
Maddison’s long-term data on labour productivity show that the major deviation from an
otherwise stationary series is the high level of labour productivity growth between 1945 and
1960. The subsequent fall of labour productivity growth relative to the preceding period, is
(7) For example, in the United States, despite the fact that the bureau of labour statistics has recently extended the
coverage of its productivity by sector index, only 47 % of employment in the services producing sector is
covered while the coverage rate in manufacturing is 100 %. Within services, the coverage also varies
considerably — 100 % in retail, 19 % in finance and 2 % in wholesale (Duke and Usher, 1998).
57
The job creation potential of the service sector in Europe
indisputable, but again this is not the case in the longer perspective. The basic pattern in the
United States is that of average productivity growth of just under 3 % during the Golden Age
which fell to around 1 % in the early 1970s and remained roughly at that level until the mid
1990s when it began to increase again.
Moreover, as was pointed out by Griliches (1994) (see introduction to this section), a possible
candidate for explaining the productivity slowdown relative to the period before 1960 may be
attributable to the increased measurement problems due to the rise of the service sector.
Indeed, in November 1999, the United States released new statistics that have adjusted the
empirical history of United States productivity growth. The new figures mean an average
annual increase in productivity since 1959 of 0.2 percentage units. The post-1995 figures are
revised upwards by an average of 0.4 %. This means that the current United States
productivity figures are close to 90 % of the Golden Age average. The adjustment is due to the
use of new data previously not available, the reclassification of computer software (from
business expense to investment) and improvements in the methodology for estimating
inflation and productivity (a consistent geometric mean technique). These revisions have had
major impact on services. For example, a new index of output attempts to incorporate the
productivity gains from the introduction of new technology in the banking sector. Despite the
fact that the new classification of computer software has received most publicity, the major
source of the upward revision is due to the other measurement improvements.
A major issue in the new growth and productivity literature is the issue of whether the
advanced economies are converging or not. The empirical evidence is mixed, see Temple
(1999). While there is evidence of convergence among OECD countries, this is of course a
somewhat biased sample as a degree of economic development is required for membership.
Of particular interest in this context is the issue of whether changes in aggregate productivity
growth are due to changes in sector composition or within sector growth. Also of interest is
the role of these two factors for the convergence towards the United States level. Bernard and
Jones (1996) perform such a study for 14 OECD countries between 1970 and 1987 using the
OECD STAN database. They find strong evidence of aggregate convergence in total factor
productivity (TFP) but that convergence varied greatly between sectors. Manufacturing TFP
is both higher and has increased more rapidly than services but shows little convergence over
the period. Thus the observed aggregate convergence is due to the development of other
sectors, and in particular utilities. Table 1.12 presents the contribution of within-sector TFP
growth (growth effect) and changing sector composition (share effect) to aggregate TFP
growth.
58
Service employment, productivity and growth
Table 1.12. Sources of aggregate growth in total factor productivity
Growth %
Share %
Total %
Agriculture
Mining
Manufacturing
Services
Utilities
Construction
8
0
47
38
6
–3
–8
–1
– 23
40
1
–4
–1
–1
25
78
6
–7
Total effect
96
4
100
Source: Bernard and Jones (1996).
The decomposition shows that aggregate TFP is dominated by within sector productivity
growth and that due to its size, the service sector is by far the most important contributor to
productivity growth.
We are also interested in the role of sector contribution in the catch-up to the United States.
As apparent from Table 1.13, the single most important factor in aggregate convergence is the
growth in within-service-sector productivity, which contributes more than one third of
aggregate convergence.
Table 1.13. Sources of convergence of total factor productivity
Agriculture
Mining
Manufacturing
Services
Utilities
Construction
Total effect
Growth %
Share %
Total %
6
12
17
34
5
8
82
–5
4
4
16
2
–3
17
1
16
20
49
7
5
100
Source: Bernard and Jones (1996).
We conclude this section on productivity in services by presenting results of research that have
studied productivity or efficiency in service sectors that can provide a reasonable degree of
international compatibility.
59
The job creation potential of the service sector in Europe
Table 1.14. Productivity and efficiency in selected service industries
(1)
(2a)
(2b)
(3)
(4)
(5a)
(5b)
(6)
(7)
Belgium
Denmark
Germany
Greece
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
Finland
United Kingdom
Japan
3.2
3.3
2.2
2.5
3.3
3.8
n.a.
1.6
n.a.
3.1
1.8
1.2
3.1
2.2
6.3
105.0
86.6
78.5
37.1
77.6
96.6
68.7
95.3
101.3
95.2
86.8
45.4
56.4
59.5
60.3
94.1
68.6
100.7
62.2
45.7
94.8
60.3
72.3
130.1
54.8
73.4
52.8
85.9
77.6
70.7
90.1
65.8
75.5
n.a.
86.8
83.3
n.a.
84.2
61.6
69.9
99.4
38.5
94.4
62.0
79.7
1.04
1.00
0.71
0.47
0.66
0.88
1.46
0.72
n.a.
0.48
1.08
0.83
0.44
0.43
0.84
58.6
53.4
63.1
36.9
74.2
68.3
52.7
89.3
132.7
88.0
77.9
58.9
48.0
68.9
80.6
43
58
44
44
40
52
31
41
61
49
44
31
54
45
46
0.60
0.73
0.48
0.38
n.a.
0.72
0.36
0.72
0.79
0.92
n.a.
n.a.
0.19
0.85
0.80
0.63
0.52
0.62
0.56
0.65
0.73
0.73
0.64
0.56
0.79
0.59
0.69
0.65
0.75
n.a.
United States
8.2
100
100
100
0.45
100
56
n.a.
n.a.
(1)
(2a)
(2b)
(3)
(4)
(5a)
(5b)
(6)
(7)
Electricity: Gigawatt hour per person engaged, 1993.
Distribution: Distribution GDP per person engaged, 1990.
Distribution: Retail sales per employee, 1990.
Construction: GDP per person engaged, 1990.
Airlines: Operating expense per available tonne-kilometre, 1993.
Telecom.: Revenue per employee, 1992.
Telecom.: Mainlines per 100 inhabitants, 1992.
Postal services: Average technical efficiency, 1975–88.
Railways: Average technical efficiency, 1986–88.
Sources: All from OECD, except (6) Perelman and Pestieau (1994) and (7) Pestieau (1993).
The most striking feature of the results presented in Tables 1.14 and 1.15 is the generally
higher productivity figures in the United States.
Table 1.15. Case studies of service productivity
Productivity index
Germany
Spain
France
Italy
Sweden
Japan
United States
Banking 1992
Retail 1990
Construction 1990
55
n.a.
50
25
66
n.a.
100
89
73
87
n.a.
84
44
100
91
84
93
91
77
66
100
Source: Banking 1992 McKinsey (1995); Retail 1990 McKinsey (1995): Construction 1990
McKinsey (1994).
60
Service employment, productivity and growth
1.4.4. Services as a source of economic growth
In the literature of the 1940s and 1950s, the progression from an agrarian, to an industrial and
finally to a service economy was largely viewed as being the result of economic growth. The
shedding of first agricultural and then industrial labour was driven by productivity growth and
the progressively higher income elasticities in the primary, secondary and tertiary sectors. In
particular, the association of the structural shifts with increasing per capita income seemed to
point to the view that it was growth that caused the structural transformation towards the
tertiary or service sector. It is obvious that one should also consider the influence of sector size
on economic growth.
The influential growth model of Solow (1957) based on neoclassical production function
found that a most observed gross domestic product (GDP) growth could not be accounted for
by the growth of capital and labour inputs. The unexplained residual was attributed to
technological progress. Papers by Romer (1986 and 1900) and Lucas (1988) led to a
reawakening of interest in the determinants of economic growth. The new growth theory
attempted to endogenise the sources of the growth residual. Technological progress and
human capital accumulation were the two most obvious candidates. The empirical literature
spawned from this research is summarised in Temple (1999).
It is not possible to clearly summarise which broad sectors of services are conducive to
economic growth. There are numerous studies that find positive or no effects of say,
education, finance or health services on economic growth. The observed effects will
obviously vary from case to case and from time to time. I briefly and somewhat precariously
outline my impression of the empirical literature.
Empirical research shows that the private returns to research and development are high. As
there are presumably considerable externalities in research and development, one should
expect the social returns to be even higher. While research on the contribution of finance is
controversial and less conclusive, the balance of evidence seems to suggest that it does appear
to have significant positive effects on economic growth (Levine, 1997). There is strong
evidence of the growth-promoting effects of health services in less-developed countries, but
less evidence in more-developed countries. This research claims to be able to account for the
reverse causality. There are also strong theoretical grounds for supposing that educational
services should be conducive to economic growth. The empirical evidence is rather mixed,
however, see, for example, Aston and Green (1996). Research in the importance of
government infrastructure investment in telephone and electricity capacity, has mainly been
conducted in the developing countries. I am not aware of any research on the impact of
personal or retail service on growth, and indeed it is far from obvious why there should be an
impact.
Finally, as was shown earlier in this chapter, it is now social or community services that
constitute the largest share of service employment. In most countries the majority of these
services are in the public sector. There is great variation in the size of the public sector.
Government employment as a percentage of total employment ranges from 14 % in the United
61
The job creation potential of the service sector in Europe
States, to 30 % in Sweden. The main sources of this employment are in education, and health
and social services. A question posed in the economic literature is whether the size of the
public sector has any effect on the rate of growth. The empirical results are, as usual,
somewhat mixed. Reviews of this literature, Atkinson (1985), Slemrod (1995), Agell et al.
(1997) show, however, that there appears to be no statistically significant effect. Agell et al.
term their conclusions ‘agnostic.’ Indeed one wonders whether one could ever expect to find
any effect using cross-country regressions, given that both the financing of the public sector
and the allocation of its expenditure are probably of more importance than its size.
1.5. Conclusions
Discussion on the potential of job creation in the service sector in the EU often focuses on the
higher growth of the aggregate employment rate and the larger size of the service sector in the
United States. However, while one can also observe a higher growth of the United States
service employment rate, exactly half of the widening of the aggregate EU–US employment
rate gap since the mid 1980s is due to a more rapid fall in European agricultural and industrial
employment. Indeed, in relative terms, the most striking difference in the changes of
employment structure in the United States and the EU during this period, has been the
continuation of the long-term decline in agriculture in the EU, and particularly in the less
developed EU economies. In the United States, the shedding of agricultural labour appears to
have run its course.
Thus, compared with the United States, the poor EU employment record is across all three
major economic sectors. Indeed, it is important to note that relatively speaking European
services have performed well. This observation should perhaps lead one to look for nonservice sector specific factors in order to explain some of the poor relative EU performance.
This is not to say that the basic premise of this report, that there is a potential for job creation
in the service sector, is incorrect. Job creation will almost certainly continue to be largely in
services. However, comparisons with other sectors of the economy, and in particular some
consideration of the European experience of the simultaneous decline in industry and
agriculture with an increase in services, should focus analysis and policy on the process of
structural change.
While one should always be wary of predicting future developments by a simple extrapolation
of current trends, many factors suggest that employment in agriculture (and to a lesser extent
industry) will continue to decline and that it will tend to be re-allocated in the service sector.
In the more developed nation States the shedding of labour out of agriculture and industry to
services is a very long-term trend and there is little reason to suppose that it will be reversed.
In the less developed States (in both Europe and elsewhere), the long-term trend has been
somewhat different in that the largest employment sector has shifted directly from agriculture
to services without an intervening period of industrial dominance. Indeed, one could argue
that one may expect the shedding of agricultural labour from the less developed Member
States to accelerate in the future. The key issue here is, of course, the development of the
62
Service employment, productivity and growth
common agricultural policy. Despite acute political difficulties, the pressure for reform,
namely, a lowering of protection and thus employment, is intense.
This chapter has shown that there has been a continuous process of EU catch-up to United
States levels of service employment share. While the catch-up has generally been faster in the
less developed Member States, it is still these economies that account for most of the EU’s
service share gap with the United States. This further underlines the relevance of viewing the
potential for service job creation in Europe in terms of structural shifts from, primarily,
agriculture to services.
Producer services are perhaps the most interesting of all the major service sector
subcategories. It is the source of the largest aggregate EU gap to the United States and,
although the gap is largest in the less developed Member States, it can be found in all Member
States except Luxembourg. This sector includes many of the jobs that are typically associated
with the rise of new technology and work organisation. While the EU deficit in this sector may
be viewed with concern, it is noteworthy that the EU catch-up with the United States in
producer services is proceeding at a faster pace than in all service subsectors (both in terms of
employment share and rate).
The second largest gap is to be found in personal services. This can mainly be ascribed to the
larger hotel and restaurant sector in the United States. While the gap is present in every
Member State except Spain, the gap is considerably smaller in the less developed Member
States, namely, the opposite of the European distribution in producer services. The European
gap in this sector should not be viewed with the same degree of concern as for producer
services as one would be less likely to find the type of productivity enhancing spillover to
other sectors that could be expected in many producer services.
The distributive services gap is appreciably smaller than that in producer and personal
services. The major subsector accounting for this gap is trade and repair. This sector exhibits
a similar distribution among the Member States as hotel and restaurants, and indeed Italy,
Spain and Greece have a larger trade and repair sector than the United States. Again the
productivity and growth effects of these services is not likely to be very significant.
In the largest service sector, social services, the EU average is at the same level as the United
States figure. However, there is a great variation within the EU, particularly in health and
social services. Also, the education sector is smaller in every Member State except Belgium.
While job quality in this sector is rather heterogeneous, the European shortfall in health and
education may be viewed as problematic, given that growth theory suggests that these sectors
may be very conducive to economic growth, particularly among less developed economies.
On the other hand, the United States health service is notoriously inefficient.
The realisation that European agricultural and industrial employment is declining as service
employment is increasing leads one to place policy for job creation in the service sector in the
context of structural change. The classic policy prescriptions for the promotion of structural
change focus on labour mobility, both occupational and locational. Thus, active labour market
63
The job creation potential of the service sector in Europe
policy along the lines of the measures considered in the employment guidelines are
appropriate. The guidelines do not, however, explicitly set labour market policy as part of a
model of structural change. As agricultural (and some older traditional industrial) activities are
intrinsically regionally concentrated, geographical mobility should be a particularly important
priority. Moreover, the case for focus on geographical mobility policy is further strengthened
by the advent of the single currency. According to economic theory of monetary unions, the
possibility of asymmetric macroeconomic shocks, together with the loss of the devaluation
prerogative, makes the primary labour market policy implication of the single currency the
promotion of geographical mobility.
Legislation on freedom of movement has a long history and is a policy area that, for good
reasons, the European Union has a strong mandate. The Treaty of Rome dealt with the right to
mobility, the single market programme made some progress in enabling mobility
(qualifications and social security) and recent initiatives have attempted to deal with
information problems. Perhaps it is time to go a step further with measures to actively promote
labour mobility. Some Member States currently provide both cash mobility grants and moving
cost assistance within their national borders. Other policy options such as dwelling quotas
may also be feasible. It is certainly a policy perspective that should be explored further.
Another broad policy perspective is related to the size of the market. Economies of scale and
the division of labour are a prime motor for economic growth. In particular, recent literature
in industrial economics places much emphasis on the role of economies of scale and scope in
generating producer services. To the extent that these jobs are driven by economies of scale,
the lack of a fully developed single market for services is an obvious candidate for explaining
the European shortfall in services. The policy prescriptions are largely along the lines of the
1985 White Paper on completing the internal market, that is, those concerning the right of
establishment and other matters of company law and occupational qualifications.
64
Chapter 2
Unbalanced growth and employment
in services
Leonello Tronti, Roberta Sestini and Andrea Toma
2.1. Introduction: The Baumol model
Cross-country research on employment trends shows undoubtedly that employment growth
during the last two decades has been slower in European countries than in the United States.
Moreover, it is unanimously accepted that the development process of industrialised countries
has moved from the ‘industrial model,’ characterised by a positive correlation between
productivity growth and employment growth, to a ‘post-industrial model’, where the
correlation is negative (Appelbaum and Schettkat, 1995). Nowadays, in fact, employment
seems to grow for the most part in industries where productivity increases at quite slow rates,
like most services. This pattern of employment growth is able to explain at least part of the
better performance of the United States. As Europe is lacking a substantial proportion of
employment in services, EU policy-makers face the problem of devising measures for fully
exploiting the job creation potential of this sector.
Having in mind this policy purpose, our analysis intends to explore the reasons for employment
expansion in services, stemming from a fundamental theoretical framework, represented by the
Baumol model of ‘unbalanced growth’. In his seminal article (1967), Baumol proceeds on the
hypothesis that economic activities can be grouped into two main sectors: technologically
progressive industries (P), with productivity growing at a constant rate r, and non-progressive
ones (NP) where, for the sake of simplicity, labour productivity is assumed to be constant. The
model is particularly interesting for studying the behaviour of employment in the service sector
as service industries are generally characterised by slow productivity growth, while agriculture
and manufacturing show sustained long-term productivity increases.
Assuming that labour is the only factor of production, and denoting with YNP and with YP the
values of outputs respectively in the NP and in the P sectors, the production functions for the
two sectors may be written as follows:
65
The job creation potential of the service sector in Europe
YNPt = a LNPt
YPt = b LPt ert
(1)
(2)
where LNPt and LPt are the amounts of labour employed in the two sectors, and a and b are
constants.
Baumol calculates that due to some degree of mobility between the two sectors, in the long
run, wages in the NP sector cannot lag behind those in the P sector. In order to simplify the
model, wages are taken to be equal and to grow at the same rate, although the results stemming
from the analysis do not vary if one allows for some divergence in the wage levels and in their
movements. Therefore, the wage level in both sectors is fixed at Wt and grows at the same rate
as productivity in the progressive sector:
Wt = W ert
(3)
where W is a constant.
This simple model shows that, in a closed economy, if wages grow at the same rate in both
sectors of the economy (in particular at the rate of growth of productivity in the progressive
sector, assumed to be constant), there will be relevant, though asymmetrical effects on prices.
In particular, costs and prices in the slow productivity sector will tend to rise without limit
relative to those in the progressive sector. In fact, denoting with Ci the cost per unit of output
in each sector, we obtain:
CNP = Wt LNPt / YNPt = W ert / a
CP = Wt LPt / YPt = W / b
(4)
(5)
It should be pointed out that the relative costs would exhibit this pattern quite independently
of the hypothesis about the behaviour of wages in (3). It is in fact easy to show, by simply
looking at (4) and (5), that
CNP / CP = (WNP/WP) * (b ert / a)
(6)
tends to infinity, irrespective of any actual value that the ratio of NP to P wages can reach.
As a consequence, in a model of unbalanced productivity growth, the survival of the NP sector
relies upon the price elasticity of demand for its output: if demand is not highly inelastic, that
is if it reacts to the continuing ‘cost disease’ of the NP sector, increases in relative prices would
lead to slow employment growth and even to the dying out of non-progressive activities (1).
However, this is hardly the general case with the whole service sector; as a matter of fact,
employment in the NP sector may expand, despite rising relative costs, if the demand for
(1) More generally, if the prices of non-progressive services increase with respect to the prices of progressive
manufactured goods, consumers should tend to substitute services with durable goods and/or self-provided
activities. This is what has happened with many personal and family services, like barbers, tailors and clothing
repair services, shoeshine services, umbrella repair, etc.
66
Unbalanced growth and employment in services
services is highly price-inelastic or is income-elastic (if goods and services produced by the
NP sector are Engel-superior), if the sector is subsidised or, furthermore, if NP services are
provided by the public sector.
If, on the contrary, the output shares of the two sectors in real terms remain constant over
time (2), not only will the non-progressive sector survive, but an ever increasing proportion of
the labour force will shift to it. In fact, under the hypothesis that:
aLNPt
YNP
=
= K
YP bLPtert
(7)
and assuming that the total supply of labour L = LNP + LP is fixed, we have:
LNP = (L – LNP) K ert = L K ert / (1 + Kert)
(8)
LP = L – LNP = L / (1 + Kert)
(9)
and
It is clear that, when t tends to infinity, the whole amount of the labour force will be transferred
to the NP sector, while the amount of labour in the P sector will approach zero.
It may be useful to illustrate this result by means of a figure.
Figure 2.1. The expansion path of the economy and sectoral shares in the labour force
(2) Baumol indicates that this will happen as a consequence of government subsidies or when the demand for the
NP sector’s services is sufficiently price-inelastic or income-elastic. The typical case is the one of local
government services, but similar behaviours are shown by Engel-superior services like education, performing
arts, shows and travels, information services, etc.
67
The job creation potential of the service sector in Europe
Figure 2.1 shows two transformation curves (the thick lines), which describe all the possible
combinations of outputs in the two sectors that can be produced with a constant supply of
labour. The slope of a transformation curve (dYNP/dYP) expresses the opportunity cost of
producing an additional unit of YNP in terms of YP. The transformation curve exhibits a
movement in the direction of the arrow, as the cost of producing YNP increases with time, if
expressed in terms of the number of units of YP that must be sacrificed. Furthermore, the NP
share of the labour force can be read as the ratio of the distances OAi/OD. Under the
hypothesis that the sector shares in real output remain constant, the ray OBC represents the
expansion path of the economy. As time tends to infinity, the transformation curve tends to
become parallel to the abscissa, and the whole labour force enters the NP sector. This result
implies a long-term shift of employment toward the NP service industries, under the
assumption of a fixed supply of labour.
Finally, the model makes a prediction about the overall rate of growth of output. In particular,
as the relative prices of the NP sector grow (12), its real output constant share will continually
increase in money terms, so that the relative output of the progressive sector will conversely
shrink, bringing about a decline in the economy’s aggregate productivity growth rate and
eventually leading the economy to asymptotic stagnation. This result is easily shown, taking
as an index of aggregate ratio a weighted average of outputs in the two sectors:
I = BNP YNP + BP YP = BNP a LNP + BP b LP ert
(10)
Substituting for YNP, YP from (8) and (9), it comes out that:
I = L (K BNP a + BP b) ert / (1 + K ert) = R ert / (1 + K ert)
(11)
where R = L (K BNP a + BP b). Hence, the percentage rate of growth of output is given by:
(dI/dt)/I = r / (1 + Kert)
(12)
confirming the intuition above, according to which the model foresees a rate of growth for the
whole economy which tends asymptotically to zero as t increases.
2.2. The predictions of the model
To summarise the results of the Baumol model presented above, there are three fundamental,
explicit predictions:
(1) Endogenous inflation or ‘cost disease’: The relative cost per unit of output in NP
industries will tend to infinity. This outcome, that may not show in aggregate price change
estimates, is the result of sector prices balancing each other: the non-progressive services’
inflationary bias can be more or less compensated for by progressive manufacturing and
agricultural deflationary behaviours.
(2) Relative long-term decline of the (agricultural and industrial) progressive sectors: The
decline should occur both in terms of employment share and output money share. Such
68
Unbalanced growth and employment in services
long-term trends can be looked at through the lenses of ‘tertiarisation’ or
‘deindustrialisation’ processes.
(3) Asymptotic stagnation: This third event is the result of a long-term slowdown both in
aggregate productivity growth (in money but also in real terms), and in output growth, that
copes with the structural shift of money and men from the progressive to the nonprogressive sector. Quite obviously, the fall in output can be compensated for by a
significant increase in the level of employment.
Moreover, even if the model does not allow for any explicit prediction on employment growth,
it is further characterised by two additional, implicit predictions concerning the labour market:
(4) If the assumption of a constant labour supply is relaxed, employment growth will vary
through time, according to the size of sectoral shares in employment. As the nonprogressive share grows, the average productivity will slow down, but the employment
content of economic growth, as expressed by the output elasticity of employment, will
probably increase. At each stage of growth, then, the overall employment outcome will
depend on the composition of the two opposite effects.
However, by definition, employment growth equals the difference between the rates of growth
of output and productivity. Consequently, the size of employment growth depends on the fact
that the long-term slowing down of productivity does not reflect on real output — a condition
that is simply impossible in a context of fixed labour supply. In other words, in the case of
constant labour supply, the prediction of asymptotic stagnation will affect both output and
productivity while, if we assume a growing workforce, it will apply to productivity only,
output growth being increasingly correlated to employment growth (see also prediction 3).
This prediction suggests that, in an unbalanced productivity growth economy, the only way
for aggregate output growth to overcome aggregate productivity growth is the development of
employment in the NP sector; a conclusion that supports the policy efforts made in the
direction of fostering the service sector growth.
(5) The generation of an intersectoral job flow out of the progressive sector and into the nonprogressive: This flow causes an employment reallocation that usually takes place
through intergenerational job redistribution, but sometimes requires some kind of more
problematic, individual retraining. The need for retraining could affect the average
duration of unemployment spells, as low-skilled workers displaced at older ages are
expected to require longer periods and more intensive employment services in order to
perfect job transitions. Another possible outcome can be an increase in early retirement
and early withdrawals from the labour force. All these processes obviously require the
development of accompanying policies (counselling, training and retraining, etc.) as well.
2.3. Relaxing the assumptions and discussing the model
Although there is some empirical evidence about the main stylised facts set forth in the
Baumol model (see, among others, Baumol, Blackman and Wolf 1992; for Italy, Tronti and
69
The job creation potential of the service sector in Europe
Cucchiarelli 1992), fortunately its uncomfortable predictions cannot be considered to be
prevalently true, at least up to now. It is certainly true that productivity (and also output)
growth slowed down in all industrialised countries after 1973; and it is also true that after 1960
employment growth was mainly concentrated in the slow productivity service sector.
However, if one looks to the different performances of individual economies, the partial
validity of the Baumol model predictions reveals itself as marked by high cross-country
differences.
The model, though, is very simple and robust in its functioning, so that if real economies
behave differently than predicted it is possible that its assumptions are not as necessary and
robust as its logical structure. We can also reconsider this last statement and reformulate it in
a different, policy-oriented perspective: it can be of great interest to identify which countries
better succeeded in escaping what we can call the ‘Baumol trap’ and how they did it, or what
assumptions of the model they succeeded in relaxing. In the following pages we concentrate
upon four major assumptions, the relaxation of which can prevent the uneasy perspective of
asymptotic stagnation and help foster employment growth:
• Fixed labour supply.
• Long period invariance of the progressive and non-progressive sector shares in total output
at constant prices.
• The absence or irrelevance of international trade (the assumption of a closed economy).
• A uniform rate of growth of money wages in both sectors, irrespective of productivity
differentials.
The relevance of wage differentiation for employment and output growth is probably the point
that up to now has raised the widest consideration (3), but the relaxation of these assumptions
can bring interesting changes in the predictions of the model. We have already briefly treated
the effects of changes in labour supply. Furthermore, if the employment share of the P sector
should expand (or at least not shrink through time, for endogenous reasons), this could suffice
to offset the long-term tendency to aggregate stagnation (even though it would lower the
employment content of growth). On the other hand, the introduction of foreign trade into the
model could itself influence the composition of output, thus changing the P sector share.
Finally, a closer correlation between sectoral productivity and wage growth can (at least
partially and temporarily) defer the cost disease as well as aggregate productivity slowdown,
in money terms.
The following section presents a synthetic discussion of the four key elements of the ‘Baumol
trap’. We will try to derive some policy conclusions regarding the issue of the job creation
potential of the service sector.
(3) However, it should be stressed that, through the lenses of the Baumol model, the real issue is whether wages
differentiate along with productivity growth, not if they differentiate as such. Furthermore, as we observed
earlier (see eq. 6 above), the positive effects of wage differentiation along productivity growth will have only
short-term positive effects on the economy and will most likely be at the expense of very high social costs (on
this last point with reference to the American case, see Phelps 1997).
70
Unbalanced growth and employment in services
2.3.1. Constant labour pool
The question of the effects of population and employment growth on economic development
is traditionally widely discussed; its theoretical implications are quite straightforward. From
the point of view of production, an increase in aggregate output can always be traced to the
contributions of a change in labour productivity on the one side and in the size of employment
on the other. This simple identity relationship, though, says very little about how the change
affects people’s welfare. We need at least to know if output growth results in an increased
quality of average living conditions –— an outcome that, in turn, is generally correlated with
per capita income growth and/or a higher employment rate, rather than with absolute increases
in output and employment.
The Baumol model that combines the assumption of a fixed labour supply with the prediction
of long-term output asymptotic stagnation, implies a parallel stagnation in per capita
income (4). If the constant labour pool assumption is relaxed, the prediction of output
stagnation falls, but that of stagnating per capita income remains, though it may be
temporarily delayed by a growing employment rate.
Table 2.1. Real GDP, employment and labour productivity growth in the EU, the United States of
America and Japan, 1960–98
EU-15
GDP
A. Average yearly
rates of change
1960-72
1973-81
1982-91
1992-97
1960-97
B. Percentage
contributions
tgdp growth*
1960-72
1973-81
1982-91
1992-97
1960-97
United States
Total
Labour
employment productivity
GDP
Japan
Total
Labour
GDP
employment productivity
Total
Labour
employment productivity
4.7
2.5
2.5
1.7
3.1
0.3
0.2
0.6
– 0.2
0.3
4.4
2.2
1.9
2.0
2.8
4.2
2.8
2.6
3.0
3.2
1.7
2.2
1.6
1.9
1.8
2.5
0.6
0.9
1.2
1.4
9.5
3.9
4.1
1.6
5.3
1.3
0.8
1.0
0.6
1.0
8.2
3.1
3.0
1.1
4.3
—
—
—
—
—
5.5
9.7
24.7
– 13.8
8.8
95.0
88.7
76.1
114.7
91.4
—
—
—
—
—
40.8
77.0
63.0
62.2
56.4
58.6
22.7
35.3
38.7
43.0
—
—
—
—
—
13.4
20.6
25.8
34.9
18.4
85.7
78.8
73.7
67.0
81.1
Source: Based on Eurostat data.
* The two contributions may not add up to 100 % as the decomposition excludes the third interaction
term, which is almost always negligible.
(4) The assumption of a fixed labour supply can be easily converted into a fixed employment rate for a given
population size.
71
The job creation potential of the service sector in Europe
If we look at what happened to GDP, employment and productivity growth in the aggregate of
the countries that today belong to the European Union, and we compare this with what
happened in the United States and in Japan (Table 2.1), we immediately see that in the long
period from 1960 to 1998 GDP grew at approximately the same average yearly rate in Europe
and in the United States (3.1 % in the EU vis-à-vis 3.2 % in the United States), while in Japan
growth was much stronger (5.3 %). However, if we consider these aggregate results in terms
of the different contributions provided by productivity and employment, the picture is totally
different. In the European case, the contribution of employment growth was very modest (less
than 9 %), so that GDP was almost totally driven by labour productivity, that moved at an
average speed double that in the United States (2.8 % per year against 1.4 %). Also in Japan,
the strong long-term growth of productivity (4.3 % per year) accounted for more than 80 % of
GDP growth, while in the United States, economic development was fundamentally based on
a dramatic increase in the size of employment (1.8 % per year), that accounted for more than
56 % of GDP growth.
If we look at the very different achievements obtained by the United States economy on one
side and Europe and Japan on the other through the perspective of the Baumol model, we first
notice that the United States succeeded in relaxing the fixed labour pool assumption far better
than Japan and, above all, Europe. Nonetheless, the ‘Baumol trap’ appears to have affected the
Japanese and American economies more profoundly than the European economies, by
slowing down productivity growth.
Figure 2.2. Labour productivity growth rates in EU-15, the United States and Japan, 1961–97
We can, in fact, first consider the rates of labour productivity increase as the figures of GDP
growth, once we clear it from the effect of increased employment. In this case, we see that the
long-term record of the United States was quite modest, as it reached only one half of the
European growth and less than one third of the Japanese. Then, if we look at the changes of
72
Unbalanced growth and employment in services
labour productivity through time (Fig. 2.2), we notice that only Japan went through a
continuing decline in the rate of productivity growth (from more than 8 % per year in 1960–73
to 1.1 % in 1992–97). The 15 members of the EU after 1982 stabilised at around 2 % per year
and the United States, though falling during the oil crisis to a very modest 0.6 % per year,
succeeded in accelerating productivity growth thereafter, reaching 1.2 % per year after 1992.
Table 2.2. GDP, employment and labour productivity growth in six European countries, 1960–98
Germany
GDP
A. Average yearly
rates of change
1960-72
1973-81
1982-91
1992-97
1960-97
B. Percentage
contributions
tgdp growth
1960-72
1973-81
1982-91
1992-97
1960-97
France
Total
Labour
employment productivity
GDP
Total
Labour
GDP
employment productivity
Total
Labour
employment productivity
4.3
2.2
2.7
1.4
2.9
0.2
0.1
0.8
– 0.7
0.2
4.1
2.1
1.9
2.1
2.7
5.4
2.8
2.3
1.7
3.3
0.6
0.4
0.3
0.1
0.4
4.8
2.4
2.0
1.6
2.9
5.1
3.5
2.3
1.2
3.3
– 0.4
1.2
0.7
– 0.5
0.2
5.6
2.3
1.7
1.8
3.1
—
—
—
—
—
5.2
3.6
29.7
– 47.4
6.2
94.1
96.8
70.0
150.3
93.7
—
—
—
—
—
11.8
13.6
12.8
5.7
11.8
88.0
86.6
86.8
92.7
88.0
—
—
—
—
—
– 8.5
33.8
28.7
– 43.3
6.6
108.9
65.2
71.3
144.0
93.2
United Kingdom
GDP
A. Average yearly
rates of change
1960-72
1973-81
1982-91
1992-97
1960-97
B. Percentage
contributions
tgdp growth
1960-72
1973-81
1982-91
1992-97
1960-97
Italy
Total
Labour
employment productivity
2.9
1.4
2.7
2.5
2.4
0.2
0.0
0.6
0.5
0.3
2.7
1.4
2.0
2.0
2.1
—
—
—
—
—
5.6
0.7
23.0
21.0
13.0
94.2
98.5
76.7
77.6
86.5
Sweden
GDP
4.1
1.8
1.9
1.4
2.5
Spain
Total
Labour
GDP
employment productivity
Total
Labour
employment productivity
0.6
0.9
0.5
– 1.4
0.3
3.5
1.0
1.3
2.8
2.2
7.2
2.5
3.3
2.0
4.1
0.6
– 1.1
1.2
0.6
0.3
6.6
3.6
2.1
1.4
3.7
—
13.9
—
47.2
—
28.7
— – 98.7
—
11.1
85.5
52.9
70.3
201.0
88.5
—
—
—
—
—
8.1
– 43.9
36.4
30.1
8.5
91.4
147.1
63.2
70.7
91.6
Source: Based on Eurostat data.
73
The job creation potential of the service sector in Europe
Within Europe, the long-run relationships among GDP, labour productivity and employment
growth are much more homogeneous. In particular, if we look at Germany, Spain, France,
Italy, Sweden and the United Kingdom (Table 2.2), we can see that the long-run GDP
performance ranged between around 2.5 % per year (Sweden and Britain), and 4 % per year
(Spain). The employment record was even more homogeneous, ranging from 0.2 % per year
(Germany and Italy) to 0.4 % per year (France). Consequently, the contribution of productivity
to GDP growth was always very high (from 86 to 94 %), and the average productivity gains
were considerable (from 2.1 % per year for the United Kingdom to 3.7 % per year for Spain).
On the other hand, the productivity slowdown was felt particularly heavily by Spain and Italy,
and that in the pre-oil-crisis period could benefit from catching-up effects, while Sweden
alone appears to have entered a path of accelerating productivity since 1982.
The simple empirical evidence presented in this section shows that, in line with Baumol’s
perspective, there could be a trade-off between employment and productivity growth; not only
because of the constraint that productivity imposes on employment once the demand is given,
but also because employment could be substantially fostered only in slow-productivity
industries, at the expense of a slowdown in aggregate productivity, even greater than the one
resulting from unbalanced growth in a fixed labour-pool context. Bearing this in mind, the
gains coming from employment growth should be weighted against the long-term effects on
per capita income.
2.3.2. Long-term invariance of sector shares
In the Baumol model, as assumed in the original paper (Baumol 1967) and later discussed and
empirically tested with reference to the United States (Baumol, Blackman and Wolf 1992), the
real output shares for the two sectors remain constant over time. The reason for such
invariance, though, is left unexplained. In order to clarify this crucial assumption, it should
first be necessary to ascertain whether this aspect has in effect the character of a ‘statistical
regularity’ taking place through time and space. The question may look simple, but it is not,
as measuring the value in real terms of service output is often a complex task (5).
If it were possible to provide reliable empirical evidence of the invariance, it would still be
necessary to formulate some (possibly testable) hypotheses for such an odd regularity that
seems to stand against continuing cases of NP; traditional products, services and occupations
being crowded out by P products, services and occupations. One possible explanation could
be traced to NP goods and services being perfect complements of P goods and services, in
production and/or in consumption.
Further insights in the direction of complementarity could come by breaking down the NP
sector looking, for instance, at who buys NP goods or services (individuals and families, firms,
the government, the rest of the world). Any one of these alternatives brings with it different
(5) The problems posed by the measurement of output and productivity in the service industries are addressed, in
this volume by the paper by Donald Storrie.
74
Unbalanced growth and employment in services
patterns for the NP sector to grow more or less proportionally to the P one. If the demand for
NP services comes from individuals and families (as in domestic work, hotels and restaurants,
tourism etc.), its growth will crucially depend upon wage differentials among P and NP sector
workers, and the complementarity will be based upon the evolution of wage differentials. If it
comes from the government (for example, schooling and research, public health services, mail
services, transportation etc.), the size of employment will depend upon the tax burden that is
tolerable for the economy, and complementarity will obviously be related to the intensity of
the ‘cost disease’. If it comes from firms (like counselling, accounting, marketing, training,
research and development services etc.), the size of demand will depend upon the capacity of
NP services to positively affect the firms’ competitiveness, both in the domestic and the
international markets (6). Finally, NP demand could come from the rest of the world (as in the
case of tourism, cultural and art services, superior education etc.), which implies that these
services are tradable. In this case, NP employment growth will depend upon the international
competitiveness of the NP sector. If we look at all these alternatives and see how single
economies have succeeded in making up their specific mix and pattern of NP sector growth,
we may reach the conclusion that long-term invariance is an unlikely, if not impossible result.
A second set of issues concerning the long-term invariance of real output shares relates to the
effective growth path of technical progress. Here the main question to answer is: how much
are production processes at the sectoral level structurally rigid in terms of the possibility of
technical progress application? In other words, is an industry whose production process is NP
today, NP forever? This question can be considered in a different, empirical perspective, by
analysing whether productivity growth at the industry level only depends upon structural
causes (the kind of output produced), or it can also be linked to other aspects, like the
evolution of technologies, market structures, the role of international trade, etc.
In order to offset this basic rigidity of the Baumol framework, it is necessary to more carefully
develop the role of demand, which is totally disregarded in the original model. According to
Inman (1985), employment growth in the service sector relies on three different factors,
namely:
• the difference between the productivity growth rates in the two sectors (weighted by a factor
representing labour intensity),
• the income and price elasticities of the demand for services,
• and, finally, eventual shifts in consumers’ preferences and in the demand for NP inputs by
firms.
The author, using United States data for the period 1929–65, shows that the increase in the
employment share in the NP sector was predominantly due to slower technical progress, or to
greater labour intensity (55 %), as predicted by the Baumol model. More recent analyses,
however, question the magnitude of the contribution stemming from technical progress. In
(6) In this case, it may be possible to observe more clearly complementarity in production.
75
The job creation potential of the service sector in Europe
addition, by introducing the hypothesis of imperfectly competitive inputs and goods markets,
it can be shown that the magnitude of the employment growth rate in the NP sector may
crucially depend on the degree of monopoly power of producers (Brunello and Scaramozzino
1992).
More recently, some studies (7) have attempted to reformulate Baumol’s model taking into
account not only the role of demand, but also the complex interactions between demand on
one side and prices and per capita income on the other. These analyses have succeeded in
throwing some light on the implications of the model for aggregate employment (8). In fact,
introducing explicit sectoral demand functions, it is not only possible to clarify the results
reached by Baumol in his seminal contribution, but also to find new ones.
In particular, the models study the role of aggregate demand in determining the size of
aggregate output and employment. At a first stage (Notarangelo 1999a), it is assumed that
demand does not influence prices, while considering the effect of prices on demand. The
production technology, as well as the behaviour of wages and of prices are consistent with
those stated by Baumol (1967) and described in Section 2.1. In addition, per capita demand
(ci) is hypothesised to determine total output according to the Keynesian principle of effective
demand:
Yit = cit Nt
i = NP,P
(13)
where Nt = N0 egt is total population, assumed to grow at a constant rate g, and per capita
demand is assumed to grow at the constant rate ρ It is clear that in an economy always at full
employment the rate of growth of sectoral demand would be equal to g + r, that is the sum of
the rates of growth of population and of productivity. However, this statement does not hold
true in an economy experiencing both unbalanced growth paths and unemployment. In order
to focus on situations where the labour market is not necessarily clear, the model distinguishes
population from total employment:
Lt = µ t Nt
(14)
where µ , representing the ratio of employment to total population, is determined by
technology and by the demand for consumption goods. Using equations (1), (2) and (13), it
comes out that:
µ = (1 / a) cNP + (1 / b) cP
(15)
µ = (L0NP / Y0NP) c0NP eρNPt + (L0P / Y0P) c0P e(ρP — r)t
(16)
or
(7) Notarangelo 1999a, 1999b.
(8) A reminder: Baumol’s model gives an insight into the possible explanations of the changes across time in the
composition of production, but it focuses exclusively on the relative allocation of labour and output.
76
Unbalanced growth and employment in services
Then, the time path for µ will depend upon the values of the parameters r and ρ. In particular,
it is possible to show that under the hypothesis of constant relative outlays in the two sectors,
the growth rate of µ is equal to ρNP or, equivalently, to ρP — r. This result is rather intuitive if
we reckon that aggregate employment is subject to two opposite forces: on the one side it is
determined by the rate of growth in the demand for goods, while, on the other, the rate of
growth of productivity may reduce the need for labour, thereby depressing overall
employment. Furthermore, depending on the relative values of r and ρP, the rate of growth of
employment may even be negative, when the growth of demand in both sectors is not able to
offset the rise of productivity in the P sector.
Further examining the behaviour of nominal output, the model reaches the conclusion that in
the short run the influence of technological progress may determine some slowdown, due to
unbalanced growth in the two sectors. However, in the long run, the rate of growth of nominal
output is simply determined by the rates of growth of both demand in the progressive sector
and population (9).
Even in the case of constant relative outputs (as stated in (7)), the results are quite different
from those obtained in Baumol (1967). In fact, if the growth rate of demand in both sectors is
smaller than the growth rate of productivity, the economy may face a period of increasing
unemployment. This will not be frictional (that is, due to the time necessary for the relocation
of labour from one sector to the other), but rather a mix of technological and involuntary
unemployment. Whilst in this model (Notarangelo 1999a), the demand for labour decreases
for a certain period of time before finally approaching the growth rate of demand, allowing for
demand to be influenced by per capita income and by income distribution bringing about nontransitory effects. The reason behind this result is that a persistent rate of unemployment may
undermine the capacity of the economy to adjust to a positive rate of growth because of the
negative influence of unemployment on per capita demand and, finally, on the aggregate
employment rate of growth.
These results differ markedly from those reached by Baumol’s model, since in the latter the
growth of output, and hence of employment, reaches a ceiling because the economy is supplyconstrained (see Section 3.1), whilst in the above analysis it is demand-constrained so that
unemployment may result.
Even more diverging results are obtained in Notarangelo (1999b), where demand in both
sectors is made dependent on prices and on per capita income. In this case not only demand
affects employment, but the employment level influences demand as well. These complex
mutual interactions bring about relevant changes in the dynamics of the system. First of all, a
demand fall (even in only one sector) may cause a decrease in the level of aggregate
employment, such that the economy is no longer able to recover. Secondly, a positive rate of
employment growth is feasible only when demand is increasing in both sectors. A noteworthy
(9) One should note that it is also possible to have a positive rate of growth of per capita real output in the presence
of a decreasing employment rate.
77
The job creation potential of the service sector in Europe
policy implication is that fluctuations in the level of aggregate employment cannot be
neglected as frictional, and call for an appropriate intervention.
These results should not be considered by themselves, but they can provide insight into the
complex issue of complementarity between goods and services in actual economies, which
can help clarify the mechanisms of job creation in the service sector.
2.3.3. The hypothesis of a small open economy
A third limitation in the Baumol model refers to the assumption of a closed economy. This
assumption is evidently at odds with the prevailing situation of industrial economies being
increasingly open to international trade, could be justified only if international trade had no
effect on the results of the model (10). However, the direct and indirect effects of international
trade on the creation of employment in the service sector may be relevant, so that it is
necessary to explore the changes induced by the relaxation of the closed economy assumption
on the basic model.
Among the various attempts to include the hypothesis of an open economy into the picture,
the more widely known is probably the so-called Scandinavian or ‘small open economy’
model (11), formulated in its simplest version by B. Balassa (1964), and later by O. Aukrust
(1977) and G. Edgren, K.-O. Faxén and C.E. Odhner (1973). According to this, it is possible
to distinguish two sectors, an ‘exposed’ sector (E) and a ‘sheltered’ sector (S), producing
respectively tradable and non-tradable goods. All goods and services subject to international
competition are included in the E sector, whilst goods and services protected from
international competition (the construction industry, government activity, housing and, to a
great extent, service industries) belong to the S sector. The assumption of a small open
economy implies that a country cannot influence world market prices, neither through its
supply of export goods nor through its demand for import goods.
In analogy with Baumol’s model (1967), in the Scandinavian model the E and S sectors differ
in that in the first sector technical progress is more pervasive, with the role of spurring on
international competition; thus labour productivity grows at a higher rate than in the S sector.
In this way, sector E tends to coincide with that of P and sector S with that of NP. Furthermore,
due to the higher degree of competition in the tradable sector, firms in sector E behave as price
takers, while prices in sector S are jointly determined by unit labour costs and a mark-up.
Again, similarly to Baumol’s model, the rate of growth of money wages in the two sectors is
the same, being determined in the E sector by the world rate of inflation along with the
productivity growth rate. Due to ‘spillover effects’ from the exposed to the sheltered sector, in
sector S wages exhibit the same rate of growth as in E, whilst the rate of growth of prices is
also influenced by the sectoral growth of productivity.
(10) Baumol explores the non-trivial relationships between productivity growth and international trade in a paper
subsequent to the one presenting the original model (Baumol 1986).
(11) The description of the Scandinavian model is based on the treatment by Frisch, 1983.
78
Unbalanced growth and employment in services
By solving the model for the inflation rate, one sees that price dynamics in a small open
economy depend on the world inflation rate as well as on the spread between the productivity
growth rates in the two sectors, weighted by the S sector’s share in total expenditure (Aukrust
— Edgren, Faxén and Odhner, or AEFO equation).
Consequently, if income elasticity of demand for the output of the S sector is large (a quite
realistic hypothesis in the case of demand for services), the economy will bear an increasing
cost for the production of services in the S sector. To summarise, if S/NP services are
complementary to E/P output, it is likely that the pattern of wages and prices in the former
sector affects overall growth and prices through international trade. Considering that a large
proportion of services belong to sector S, their cost and quality become extremely important
for the country performance on the whole, in that they can influence the exposed sector’s
productivity and output growth. For this reason, a larger share of S/NP services (in terms of
both output and demand) can be desirable and necessary to recoup a country’s
competitiveness.
It should be pointed out that the Scandinavian model relies heavily on structural forces and
aims at explaining long-run tendencies (12), irrespective of shorter term modifications.
However, the model has been extended in order to take into account better the role of demand.
Once demand in each sector is introduced (13), new insights arise. In particular two opposite
forces influence the rate of growth of demand for the S sector: on the one side the (negative)
inflation effect, given by the difference between the rates of inflation in the two sectors, and
on the other the (positive) income effect, weighted by the income elasticity of demand. Among
the main results of the Scandinavian model, and due to the assumption of constant income
shares in the two sectors, ‘the income elasticity of demand for non-traded goods must exceed
that for traded goods’, (14) as is the case with NP goods in the original Baumol model. The
intuition behind this result is clear: the potential decline in demand for non-traded goods and
services, caused by the continuous increase in the relative price of the S sector’s output, has
to be countervailed by a higher income elasticity of demand. Employment in the S sector
should grow where the services produced are Engel-superior.
A further interesting extension of the Scandinavian model, due to Branson and Myhrman
(1976), stems from the criticism of the hypothesis of invariant shares of the S and E sectors in
real output. Let us assume that the policy-maker is able to set a target for the unemployment
rate through demand management, causing an increase in both the inflation rate and the
employment rate. The adjustment mechanism involves a pressure towards an increase in the
(12) As Frisch (1983) notes, ‘the Scandinavian model is a long-run supply model with implicit demand conditions.
Real growth in both sectors is exogenously explained by the rate of growth of productivity and employment’.
(13) The S sector produces only non-traded goods, demanded by residents, while demand for goods and services in
the E sector is made up of three components: (i) the demand of domestic and, (ii), foreign residents for
domestically produced tradable goods, and, (iii), the demand of domestic residents for foreign-produced
tradable goods. Furthermore, the model assumes that the current account is always in balance, that is imports
are always equal to the difference between domestically produced tradable goods and the demand for them by
domestic residents.
(14) Frisch, 1983.
79
The job creation potential of the service sector in Europe
growth rate of wages, which is not troublesome in the S sector, where prices are fixed
according to a mark-up rule, whereas it is such in the E sector where, on the contrary, prices
are exogenously determined in the rest of the world, so that higher money wages lead to a
decline in profitability. In addition, the growth in income, along with the higher income
elasticity of S services, induces a further increase in the demand for S output. Therefore, real
growth accelerates in the S sector and decelerates in the E sector, thereby leading to a process
of accelerated shifting in output and employment shares. The results achieved within this
framework prove to be in accordance with a stylised fact of many European labour markets in
the 1980s when, in a context of moderate but persistent inflation, service employment grew
and industrial employment simultaneously declined (15).
It may be useful to mention also the stream of research that focuses on foreign trade as a
relevant factor in determining both growing wage inequality in the United States (or the rising
unemployment rate in the EU countries) and the displacement of workers from manufacturing
jobs, with adverse consequences for income distribution. As Wood (1994) suggests, the trade
of industrialised countries with less developed countries (North-South trade) consists mainly
of the exchange of more skill-intensive goods for less skill-intensive ones. It is then possible
that trade raises labour productivity, as more industrialised countries move into activities that
make more use of skilled labour, and react to competitive pressures from the South by
devising ‘defensive’ innovations that reduce the demand for unskilled labour.
The so-called ‘deindustrialisation argument’, then, blames trade as a major cause of stagnating
or declining incomes among low-skilled American workers. The argument claims that lowskilled workers face a wage differential between manufacturing (where high wage jobs are
concentrated) and other sectors: growing international competition would lead to a loss of
many of these jobs. In principle, an increase in manufacturing imports should lead to a
reduction in the country’s welfare, as some workers are pushed out of the high wage sector
into the low wage sector. However, the measured welfare loss from deindustrialisation shows
that a more interesting effect does not lie in the amount of workers displaced from the high
wage sector, but in the wage compression suffered by those who remain in that sector in the
face of a reduction in the demand for low-skilled labour. It is well known that since the 1970s
there has been a dramatic widening of wage differentials in the United States. To give an
example, real wages of workers at the 90th percentile have risen about 15 %, while those of
workers at the 10th percentile have fallen about 25 %. The widening of wage differentials has
been less dramatic in other advanced countries, where there has been a secular rise in
unemployment that is widely regarded as the result of an attempt to suppress the global
pressures for growing inequality. Over the same period, international trade (especially the
exports of manufactures from low wage countries), has increased substantially.
In principle, the growth of North-South trade could explain the rise in wage inequality in the
United States: if imports of labour-intensive goods have led to a fall in the relative prices of
(15) The case has been particularly evident in Italy, where such a structural shift has been reinforced by public
expenditures financed by rapidly growing public debt.
80
Unbalanced growth and employment in services
theses goods, it can happen that the real wages of low-skilled workers have indeed been
lowered. This argument, however, is subject to severe criticism. As Krugman (1996) suggests,
the deindustrialisation hypothesis is of little practical relevance. As a matter of fact, trade
flows are too small to explain the large change in factor prices that occurred during the last
decades in the United States, in particular, the substantial rise in the wage premium associated
with a college education (about 30 % since the beginning of the 1970s, see also Section 3.4
below). In particular, Krugman found that ‘North-South trade has lowered the relative price
of labour-intensive products by less than 1 %, and the relative wage of unskilled workers by
less than 3 %. That is, trade has caused a significant, but fairly small fraction of the massive
increase in wage inequality in advanced economies’, (Krugman, 1995). Within Krugman’s
analytical framework (that includes the fact that the American E sector is not a price taker),
some goods and services previously non-tradable become tradable as a result of globalisation.
It is then possible to draw some preliminary conclusions from the analysis of the role of
international trade on employment in services. A first observation is that market deregulation
and new technologies can both influence the exposure to international competition of
previously sheltered services. Phenomena of this kind, in turn, can spur productivity growth
in these industries, thus widening the boundaries of the P sector and consequently, delaying
asymptotic stagnation as well as potentially improving the competitiveness of the E sector
with positive effects on employment in both P and NP sectors.
A second consideration concerns the different behaviours of the United States and European
economies in the face of globalisation. If globalisation and deindustrialisation processes put a
downward pressure on the demand for low-skilled workers in the E sector, flexible wage
arrangements in the United States accommodated the change and together with increased
wage dispersion, led to a relative stability of low-skill employment in the E sector, as well as
to employment growth in the S sector. On the other hand, in the EU, where wages are more
rigid, the same pressure caused a rapid increase in the unemployment rate. This happened
because, on the opposite side of labour demand, globalisation increases the demand for skilled
labour in the E sector and in the industries that provide services to it; but here again there are
two stories; one on each side of the Atlantic. While in the United States, employment and
wages of highly-skilled workers are rising, in Europe wages are rigid, and unemployment
affects highly-skilled workers significantly.
2.3.4. Productivity, wages and technical progress
Further examining the empirical evidence relating to the Baumol model, it is fair to say that
the fourth assumption, concerning the movement of wages and productivity, has probably
proved to be the most controversial. First, it does not seem realistic that, as Baumol argues,
the wages in the two sectors of the economy go up and down together, unless this occurs over
a long period of time (16).
(16) Baumol (1967) claims that ‘in the long run there is some degree of mobility in all labour markets, and
consequently, while wages in one activity can lag behind those in another (...) we cannot expect the disparity
to continue indefinitely’.
81
The job creation potential of the service sector in Europe
Second, the assumption of homogeneous wage growth across industries is at variance with the
evidence of growing inter-industry wage differentials, particularly wide in the United States,
at least from 1975 to 1993. The normative implication of this evidence is fairly clear: as long
as wages move along with productivity differentials (as is supposed to be true in the United
States case), they can offset the perverse effects of slow productivity in terms of price
increases (17), and guarantee employment growth at the same time (though not hampering, but
accelerating the labour force shift towards the NP sector).
Substantial research has focused on the United States wage structure of male workers in the
period between 1963 and 1989, to ascertain the role of labour demand and supply in its
dramatic change (see, among others, Murphy and Welch, 1992; Juhn, Murphy and Pierce,
1993). These studies show that, although the basic pattern of average wages has not changed
over the three decades under analysis (namely, higher earnings are correlated with higher
schooling and age-wage profiles are concave), the size of wage differentials has increased
from peak to trough by a factor of two-to-one. Thus, the pattern has not been monotonic across
time, exhibiting instead a marked trend towards increased skill premiums. In addition, the data
document that wage inequality among men with the same education and experience has
increased by 30 % since 1970.
In order to explain these changes in wage profiles, first the possible role of exogenous shifts
in the age and education attainments of the labour force have been investigated, along with a
stable factor demand. The empirical analysis worked out by Murphy and Welch (1992)
strongly rejects this hypothesis for the 1980s, suggesting that demand shifts must have played
a relevant role (18). In particular, due to high levels of business-cycle unemployment
experienced by the United States during the 1980s, wages declined for younger and less
educated workers. Similarly, in Juhn, Murphy and Pierce (1993), the trend towards increased
inequality is not viewed in terms of an increased dispersion in (unobserved) abilities, but
rather as an increasing market return to skill, due in turn to changes in demand.
In summary, the explanation for this general rise in the return to skill seems to lie in significant
shifts in labour demand, both across and within industries, that favour the most skilled. In
other words, employment has shifted towards industries and occupations that demand more
skilled labour, even in the face of rising skill premia or wages. The reduction in relative wages
for the less skilled, though, did not translate into a higher employment of this segment of the
labour force: on the contrary, the raise in the American employment rate since the mid-1970s,
is to be traced to a growth of the market for skilled labour, characterised by higher
(17) It should be noted, however, that the moderating effect of wage differentials on the ‘coast disease’ may be only
temporary (see eq. 6 above).
(18) A more detailed analysis shows that a simple supply side explanation might work for the comparison of the
wage structure in the 1970s and in the 1980s. In fact, the percentage of workers with a college degree expanded
rapidly during the 1970s, while wage differentials declined. Afterwards, during the 1980s, supply growth
slowed, while wage differentials widened. Therefore, changes in the age and education composition of the
labour force may have resulted in some of the wage changes, but they cannot be the only explanation. In fact,
co-movements of prices and quantities indicate that during this period there were substantial shifts in relative
labour demand across education and age classes, something that is particularly evident after 1980.
82
Unbalanced growth and employment in services
participation and employment rates, that countered increasing unemployment and
discouragement rates for the unskilled (Phelps, 1997).
As we have already said, if increasing wage inequality represents a well-known stylised fact
for the United States, it did not appear in European countries. In fact, with the exception of the
United Kingdom where wage differentials widened over the 1980s, during the last three
decades European countries experienced a quite stable (if not declining) level of wage
dispersion, despite a remarkable rise in unemployment rates. Thus, the key factor
differentiating European countries from the United States as to job creation performance
might be represented by a higher mismatch between productivity and wage differentials. If
this is true, if the United States had presented a wage behaviour more in line with the
Baumolian assumption of homogeneous intersectoral growth, a slower pace in the creation of
jobs in the service sector and, consequently, higher unemployment rates also among skilled
labour would have been experienced. On the other hand, allowing for job creation in the slow
productivity sector through wage differentiation could provide temporary relief from the
‘Baumol trap’, as this kind of employment may not be able to generate enough resources to
increase the living standards of unskilled workers’ families.
The different employment outcomes on the two shores of the Atlantic Ocean could be seen as
two sides of the same coin, depending on the institutional settings of the labour markets. If in
flexible labour markets the pressure towards increasing inequality transforms itself into
widening wage differentials, in highly regulated labour markets it should translate into higher
unemployment, in particular at the lower end of the wage scale. Nevertheless, even if it may
be true that European countries’ labour markets are lacking wage flexibility and exhibit higher
adjustment costs, higher non-wage labour costs and more generous benefit systems, this
explanation would not tell the whole story. Beyond the empirical evidence, it may be
worthwhile to look more closely at the causes of this pressure towards increasing wage
dispersion. In particular, skill-biased technical progress, along with different wage setting
mechanisms, might explain the observed difference between the United States and Europe in
labour market trends.
A useful starting point for analysing the impact of technology on factor prices was first
introduced by Hicks (1932), which showed, in an extremely simplified model of one good
economy, how the effect of technical progress depends on its factor bias.
83
The job creation potential of the service sector in Europe
Figure 2.3. The effect of technical progress on factor prices
Figure 2.3 above illustrates a competitive economy producing a single aggregate good and
employing two inputs, skilled (S) and unskilled labour (U). The curve I represents the initial
isoquant, while the ray OE is the ratio of unskilled to skilled labour, and the slope of the w line
is the ratio of skilled to unskilled wages. The effect of technical progress can be seen in the
figure as an inward shift of the isoquant I to I’. If the relative supply of labour remains
unchanged, the resulting relative wage will be determined by the slope of I’ at the point where
it crosses OE. Depending on the bias of technical progress, there will be a different effect on
wages. In other words, if technical progress is Hicks-neutral (and the isoquant simply shifts
inwards), the relative wage will not vary. If, on the contrary, technical progress is skill-biased,
in other words the ratio of skilled to unskilled labour rises for any relative wage level, then its
effect will be to raise the skill premium, for example, to the level represented in Figure 2.3 by
the slope of the line w’.
In order to investigate more deeply the role of technical progress as the main explanation for
the rise in skill premium, other analyses (for example, Krugman, 1995) consider an economy
that produces two goods, a skill-intensive good X and a labour intensive good Y, using as
inputs both skilled and unskilled labour. The hypothesis of having two goods (or, equivalently,
two sectors) allows us to endogenise prices of goods. In addition, to simplify analysis, it is
assumed that there are fixed proportions in each sector; that is, the allocation of resources
between X and Y is determined independently of the prices for factors and goods. It follows
that Hicks-neutral technical change in a chosen sector does not affect the factor ratio in the P
sector, nor the allocation of resources and factor prices. If, on the other hand, technical
progress investing either sector is skill-biased, the outcome is a rise in the return to skill;
biased technological change raises the price of that factor toward which it is biased. Moreover,
relaxing the fixed proportion assumption does not modify the basic results (19).
Roeger and Wijkander (1999a, 1999b) introduce a quite similar analysis. The authors consider
skill-biased technical progress in a two-sector economy that employs both skilled and
(19) On the other hand, the sectoral bias of technical change has an ambiguous effect if it is there at all.
84
Unbalanced growth and employment in services
unskilled labour, where the two sectors are represented by manufacturing and services. In their
view, adopting an explicit sectoral perspective may help to explain the fall in real wage of the
unskilled, which is still puzzling in the light of the previous analysis (20). In fact, in such a
framework, a decrease in real wages for the low-skilled may come not only from decreased
productivity, but also from increased prices for goods and services. Two types of equilibria are
examined within this model: in the first, wages and prices are allowed to adjust, whereas in
the second, relative wages are not fully flexible and the equilibrium may not be market
clearing.
As expected, in the case of rigid wages, skill-biased technical progress in manufacturing
triggers a quantity adjustment with a rise in unemployment, a decreased production of
manufactured goods and an increased production of services. If, moreover, manufacturing is
the more skill-intensive sector, the expansion of services could be limited by the availability
of unskilled workers, bringing about a higher unemployment rate. On the other hand, in the
case of flexible wages, technical progress would lead to increased wage dispersion, provided
that the elasticity of substitution between the two types of labour input is larger in
manufacturing than in services (21).
The intuition behind the last result is simple; with a large elasticity of substitution in
manufacturing, a productivity shock generates a large excess demand for skilled labour, which
will be eliminated through an increase in the relative wages. Similarly, a low elasticity in the
service sector implies that demand for skilled labour there is quite insensitive to wage
changes, making a large wage increase necessary once again. Moreover, with high elasticity
of substitution in manufacturing, the wages of the skilled increase substantially, as well as the
price of manufacturing; through this causal link the utility of the unskilled is obviously
lowered. Therefore, an interesting point resulting from the study and one not fully investigated
within empirical analyses is that skill-biased technical progress may be highly non-linear with
respect to changes in both sector shares and elasticities of substitution, as different outcomes
(not only quantitatively but also qualitatively) may emerge (22).
(20) ‘Within a macroeconomic context where workers are remunerated according to marginal productivity, falling
real wages among the low skilled would imply that skill-biased technical progress does not only increase the
productivity of the skilled but also decreases marginal productivity of the low-skilled. (...) One explanation for
this phenomenon would be that skill-biased technical progress makes low-skilled workers redundant in their
current jobs and forces them to perform tasks with lower skill requirements. This argument does, however,
need a fall in the average efficiency of low-skilled workers’. (Roeger and Wijkander, 1999b). It should be
pointed out that the model assumes that wages for the skilled always adjust so as to bring the market for skilled
labour in equilibrium. When the relative wage is rigid, unskilled workers can become unemployed.
(21) Many studies claim that the elasticity of substitution between skilled and unskilled workers is indeed quite
large in manufacturing, which is the sector with the highest productivity growth. Hamermesh (1993), for
instance, suggests a value above 3 for the United States.
(22) Finally, a policy simulation exercise was formulated using data for France since the beginning of the 1970s.
The exercise, certainly not politically feasible but still useful as a benchmark, consisted of removing relative
wage rigidity and developing a welfare analysis. As a result, ‘such a change would lead to efficiency gains but
at the cost of a welfare loss for the unskilled workers’. (Roeger and Wijkander, 1999b). In addition, the
production of both goods would increase, but more for services than for manufacturing.
85
The job creation potential of the service sector in Europe
Summing up the results of our analysis on the effects of wage dispersion, we notice that in the
United States case the long-term increase in wage dispersion since 1975 correlated with a very
high rate of job creation, particularly for skilled labour, but did not result in any marked
acceleration in labour productivity growth. According to our analysis, the European
unemployment problem could be a ‘mirror image’ of the increased wage dispersion in the
United States. Needless to say, the superior performance of the United States labour market
supports the view that institutional features play an important role in sustaining wage
inequality. However, as Freeman (1995) points out, though undoubtedly successful, labour
market flexibility has not been without cost for the United States. On the one hand it has
implied a sharp rise in inequality and growing poverty, while on the other the wage reduction
suffered by the less skilled has not increased their employment or worked time. Moreover,
even if we neglect the shortcomings of an employment strategy based on wage dispersion,
allowing for wages moving with slow labour productivity in many services requires some
labour market prerequisites. Some of these would include: educational attainments of the
unemployed should be low, unions’ pressure on wages weak, and women ready to hold a
relevant share of the jobs that would be created.
In other words, the dispersion of wages by industry may involve a strong subsidiary role for
institutional settings. After all, the United States experience seems to call (at least to some
extent) for poor quality services, with little effect on tradable goods’ competitiveness, and
heavy social costs associated with employment growth. This side of the model, even if
counterbalanced by a probable wider increase in highly-skilled jobs, does not seem desirable
for European countries, as it is not compatible with the so-called ‘European social model’.
Moreover, as unemployment in these countries is characterised by a higher proportion of
highly-skilled labour, the United States model may simply appear unlikely.
2.4. Looking for alternatives
The analysis of the Baumol model presented in the preceding pages has led us to some
conclusion about job creation in services.
1. The first important result is that service employment growth could not be in itself a good
policy objective (at least if we take per capita income growth as a proxy for the increase in
average welfare and living conditions), unless such a growth takes place in industries that
contribute, directly or indirectly, to aggregate productivity growth. In other words, the
Baumol model proves to be a very useful analytical tool in that it clarifies the terms of the
trade-off between employment and productivity growth in every economic system (23).
2. A major way to escape the ‘Baumol trap’, or else to soften its constraints, could be provided
by new technologies and organisation improvements that transfer technical progress into
previously non-progressive service industries. These changes can expand the boundaries
(23) It may be important to note that this trade-off is typically determined by demand constraints, so that its optimal
solution may require some kind of effective demand management.
86
Unbalanced growth and employment in services
of the progressive sector and thus delay the asymptotic stagnation of aggregate
productivity, even if this may leave the terms of the trade-off untouched (or even worse)
between productivity and employment growth.
3. If we take into consideration the role of international trade, we first note that exposure to
it may spur productivity growth in the industries concerned and second, that sheltered
services may profoundly affect the performance of exposed industries as long as they
provide them with domestically produced inputs. For this reason, a particular effort should
be made by policy-makers to favour international competition in services as well as
increasing domestic competition in the service industries whose output is complementary
to that of the exposed industries.
4. The choice of favouring employment growth in services by differentiating wages along
with productivity growth may prove unsuccessful in terms of welfare, as it may accelerate
the perspective of productivity asymptotic stagnation even in the presence of some positive
effects on aggregate output growth. Furthermore, in a context of unbalanced productivity
growth, the possible benefits of wage differentiation are only of a short-term nature.
Our discussion leads us to devise some alternatives to the idea of escaping the ‘Baumol trap’
through wage differentiation — an idea that, as well as being short-sighted as to its beneficial
effects, has many negative social implications. As Appelbaum and Schettkat (1995) suggest,
market forces have proved unable to deliver both high productivity growth and employment
growth. Policy-makers then have a difficult decision to make: on one hand, they may expand
less efficient activities by allowing for lower wages. On the other, they may opt for enhanced
efficiency and high-quality services by supporting investment in human capital and physical
infrastructure, knowing that productivity gains may reduce employment growth for any given
rate of labour demand. Even though it is probably not viable to chose one of the two options
exclusively, it is certainly possible to make the first alternative a socially limited and
temporary step, while pursuing the second one as the European strategy for regaining
competitiveness, for modernising and expanding employment in the service sector.
Consequently, though we cannot disregard the benefits that could come to EU labour markets
from more flexible institutional arrangements (like the one called for by the Luxembourg
strategy under the pillars of employability, entrepreneurship and adaptability), fundamental
advantages may come to Europe from the development of high-quality services for the
progressive sector, which could be able to positively influence its international
competitiveness as well as its size. In the area of high-quality business services, where it is
likely that market forces alone may not succeed in generating enough demand, the European
governments and the EU institutions should favour job creation through training, incentives
and financial facilities. In this same direction, the initiatives are crucial to modernise public
administrations as well as social and State-provided or State-financed services that affect the
competitiveness of the tradable industries.
Secondly (and in strict connection with the first point), in the face of the diffusion of
information technologies and the potential globalisation of the service markets, the
assumption of long-term invariance in the output shares of the progressive and the non-
87
The job creation potential of the service sector in Europe
progressive sectors could prove unrealistic. Here an alternative strategy seems available to the
policy-maker, based on the objective of expanding the progressive sector by transforming
non-progressive into progressive activities, by means of new technologies and improved skills
and competencies. This result could be achieved by means of measures aimed at modernising
and reorganising traditional services (like information, arts and culture, research and
education, counselling and design, finance and banking, etc.) in the direction of helping them
to become (at least partially) tradable and progressive themselves. This could relax the
demand constraint of the Baumol model and create room for employment growth in both the
progressive and the non-progressive sectors. In fact, it would be possible to achieve
productivity gains in a wide range of business, consumer and household services, which may
be considered as Engel-superior, since they are highly price-inelastic and income-elastic. In
particular, information and communication technologies may be expected to bring about
employment expansion while raising real income at the same time.
A further choice would be to support massive investment in innovation, in order to recoup
technological gaps, devise new products and open up new markets for the progressive sector,
thus increasing through this channel the effective demand for high-quality services.
In sum, the critical challenge that confronts the opportunities for employment and growth in
the European Union is based on the modernisation of the whole European service sector, in
order to make it more friendly and useful not only to social, technological and economic
innovation, but also to the rest of the economic system as well as to the rest of the world. We
have seen here that the European choice in facing the ‘Baumol trap’ has been to keep to
productivity and income growth, while reducing employment. A wise strategy for developing
high-quality services along the lines of modernisation and internationalisation could maintain
the productivity record while supporting the employment potential of the European
economies.
88
Chapter 3
Service employment: A gender perspective
Dominique Anxo and Colette Fagan
3.1. Introduction
One of the most salient features and persistent trends in advanced economies is the increased
feminisation of the labour force and the related shift from the single male breadwinner
household towards dual earner households. The overall increase of female employment rates
over the last decades has been closely associated with the development of the welfare state in
general and the growth of social services in particular. Despite the significant reduction of the
gender employment gap, there are still large differences in the patterns of female labour
market integration in Europe. The extent to which these differences are related to country
disparities in the industrial structure and variations in the pace and patterns of employment
growth in the service sector is examined in the first section of this chapter (Section 3.2).
Obviously, disparities in the female employment rate among EU countries and the United
States cannot be ascribed to country differences in the magnitude of service activities alone.
Other societal aspects are also important, such as differences in social norms affecting the
household gender division of labour and also diversities in the organisation of the welfare state
and employment regimes. A typology of European countries according to the nature of welfare
states and the provision of social services can serve as a useful heuristic device to understand
the disparities in the position of men and women in the European labour market. The extent
to which these regimes have had a differentiated impact on the female employment rate and
how the development of these welfare state regimes are linked to the development of the
service sector is also addressed in the first section.
Another central issue is the relationship between the country related employment growth in
the service sector and women’s integration versus segregation by sector and occupation. High
feminisation of the labour force and high welfare public service activities have also led to a
89
The job creation potential of the service sector in Europe
high and persistent sectoral and occupational segmentation by gender. In most European
countries, the outsourcing of traditional female household activities (childcare and elderly
care) to the labour market has clearly eased women’s transition from the domestic sphere into
the labour market, but on the other hand this outsourcing has, de facto, implied a high gender
labour market segmentation and occupational segregation. The European and United States
labour markets are still characterised by a high degree of gender segregation. The implication
of sex-segregated employment and service sector restructuring for women’s labour market
integration is addressed in Section 3.3.
Information technology plays a vital role in the growth of service employment (see J. Philpott,
this volume). The new information and communication technology has already had dramatic
effects on various service sectors. Recent developments in insurance, banking and financial
services are probably the most illustrative example. An analysis of the transformation of work
and the gender consequences of the change in the employment structure connected to the
emergence of the new ‘informational paradigm’ are also addressed. One central question in
this context is the extent to which the prevailing economic and employment restructuring
trends are favourable to women’s future employment prospects and how this transformation
will affect the pattern of occupational segregation/integration. It has been argued that the rapid
diffusion of new information and communication technology may have a detrimental impact
on employment in many service areas where women are over-represented. How the trends
towards a flatter hierarchical work organisation and greater functional flexibility may
influence the gender employment structure is also examined in Section 3.3.
In recent years, great attention has been paid to the incentive structure on the labour market in
explaining the relative deficit of employment growth between Europe and the United States.
Social transfers, taxes and non-labour costs may, either separately or jointly, raise barriers and
disincentives affecting the growth of service activities, in particular personal services. It has
been argued that a high tax wedge restricts consumer demand for household related services
limiting job creation in service activities and leading to inefficiency in resource allocation
within the household and in the whole economy. These wedges may affect household
decisions both in terms of gender allocation of time and labour supply (hours worked) and
give rise to economic inefficiency. The last section of the chapter explores the theoretical
implications of policy designed to shift or outsource tasks that are largely carried out by
domestic work or through self-provisioning, mostly by women’s labour, into the service sector
of the economy. Such a development could be expected to have implications for the gender
division of labour and allocation of time within the household, for women’s labour supply and
their career opportunities, and also for the level of employment growth and overall efficiency
of the economy. Lastly, the concluding section offers some policy recommendations.
3.2. The macro societal perspective
The main focus of this section is to analyse the macro level relationship between the
organisation of welfare provision, the employment structure and the position of men and
women in the labour market. As mentioned above, the increase in women’s employment rates
90
Service employment: A gender perspective
since the 1950s has been closely associated with the growth of the service economy, including
the expansion of public sector employment produced by the development of welfare states.
The gap between women and men’s employment rates has been reduced as a result. The
common trends of increased employment in the service sector and the rising female activity
rates are obviously interrelated but the causal relationship runs in both directions. The growth
of the service sector has favoured women’s integration in the labour market and the increase
of female labour supply has rendered possible the growth of the service sector.
However, as Table 3.1 shows, there are still large differences between countries in the level
and form of women’s labour market integration (as well as within countries, for example, by
education, ethnic status or region) (Rubery et al., 1998, 1999).
Table 3.1. Employment rates for women and men of working age in the EU, by full-time
and part-time status, 1996
Men
Belgium
Denmark
Former West Germany
Former East Germany
Greece
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
EU-15
Women
Full-time
Part-time
Total
Full-time
Part-time
Total
65
72
71
68
71
60
64
64
63
63
63
73
69
58
65
70
66
2
8
3
1
2
2
3
3
2
1
12
3
2
5
6
5
4
67
81
73
69
73
62
67
67
65
64
76
76
71
62
71
75
70
32
44
36
44
35
27
37
33
32
29
17
42
48
49
40
35
34
14
23
20
12
3
5
15
9
5
7
37
17
6
9
28
28
16
46
67
55
56
39
32
52
43
36
35
55
59
54
58
68
62
50
NB: Total may vary slightly from the sum of full-time plus part-time due to rounding of decimal
places.
Source: European labour force survey, 1996.
To be more precise, most of these national differences are due to variation in the form of
labour market integration of mothers. This is illustrated in Figure 3.1, which shows that
overall employment rates for women aged 20–49 and without dependent children exceed 70 %
91
The job creation potential of the service sector in Europe
in eight of the 14 countries shown (1), and only fall below 60 % in Greece, Spain and Italy.
Most of this employment is full-time, except in the Netherlands. In contrast, motherhood is
associated with a marked reduction in employment in many, but not all, countries.
Figure 3.1. Full- and part-time employment rates of mothers and non-mothers
aged 20–49 in the EU, 1996
To what extent are these differences related to the relative size of the service sector and the
pace and pattern of employment growth in this part of the economy?
Figure 3.2 suggests there is no deterministic relationship between the size of the service sector,
measured here by the employment share of services in total employment and the employment
rates. Even though there is a positive correlation between the degree of women’s commitment
in the labour market and the employment structure, a simple regression analysis shows that
less that 20 % of the country variation in female employment rates could be ‘explained by’
variation in the magnitude of the service sector.
(1) West and East Germany are shown separately because prior to unification they had distinctly different State
policies towards the employment of mothers, the legacy of which remains today in the sharply different patterns
of maternal employment for women in these two parts of Germany. Harmonised labour force survey (LFS) data
by motherhood is unavailable for Sweden. In Sweden, mothers have a high employment rate, for example, in
1993 three quarters of Swedish mothers with a child aged 6 years or under were employed, 35 % in full-time
employment and 40 % in part-time employment (Deven et al., 1997).
92
Service employment: A gender perspective
Figure 3.2. Female employment rates and employment share in services, EU-15 and the United
States, 1997
Source: Eurostat (1999) and own calculations.
Hence, national differences in women’s employment rates are not simply due to the magnitude
of the service sector and the diversity in employment structure appears not to be a good
predictor of female integration in the labour market. The decision to participate in the labour
market and the types of jobs available to women and men are shaped by a number of societal
features, particularly labour market and welfare state institutions, as well as more broadly by
social norms as to the conditions under which it is appropriate or expected that women (and
men) will participate in market work. Nevertheless, service sector expansion is an important
part of the jigsaw.
The concentration of employment in service sectors is shown in Table 3.2. Overall, 55 % of
male and 80 % of female employment is concentrated in the service sector, and part-time
employment is even more concentrated in this part of the economy.
93
The job creation potential of the service sector in Europe
Table 3.2. The concentration of employment in services in the EU Member States, 1996
% of men’s employment
Belgium
Denmark
Former West Germany
Former East Germany
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
EU-15
% of women’s employment
Full-time
Part-time
All
Full-time
Part-time
All
59
56
50
44
52
57
49
55
64
61
52
51
50
55
57
54
84
81
74
77
67
75
70
58
79
74
61
37
62
73
86
74
60
59
51
45
52
58
50
55
65
63
52
50
51
57
59
55
81
79
75
78
79
80
77
72
92
84
74
66
79
84
80
77
92
89
82
87
85
88
89
73
92
90
80
58
91
89
91
86
85
83
78
80
80
82
80
72
92
88
76
65
80
86
85
80
Source: European labour force survey, 1996.
Between 1980 and 1996 service sector employment in the EU increased by around 19 million,
compensating for the loss of 13 million jobs in agriculture and industry. The largest area of
growth has been in the ‘care’ services (health, social services, education), various business
services and environmental activities. There has also been a significant expansion in the
number of jobs in hotels, catering, leisure and recreational services associated with the growth
in tourism and business travel. Only a small part of the growing concentration of employment
in services is attributed to the greater prevalence of part-time employment in this part of the
economy. In most countries the share of employment which is concentrated in services is only
reduced by around 2 % if measured in terms of the volume of hours worked rather than the
number of employed persons (CEC 1997c:85–97). Nevertheless, part-time work remains
heavily concentrated in services, particularly in sales and distribution and miscellaneous other
services. These service activities accounted for 70 % of the growth in part-time work between
1983–92 in the EU (Smith et al., 1998, Walwei 1998).
Public sector service employment has played a particularly important role in the development
of women’s employment, especially in the Nordic countries. Public administration and other
public services account for at least half of women’s employment in the Nordic Member States
and Belgium (Table 3.3). Overall, the public sector accounts for nearly 44 % of female
employment in the EU-15 compared with just under 22 % of men’s employment.
94
Service employment: A gender perspective
Table 3.3. The concentration of employees in the public sector in the EU Member States, 1996
% of men’s employment
Belgium
Denmark
Former East Germany
Former West Germany
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
EU-15
% of women’s employment
Full-time
Part-time
All
Full-time
Part-time
All
25
23
19
21
20
22
22
24
23
23
19
20
21
21
20
21
37
33
44
35
36
45
35
26
26
24
34
24
24
37
26
32
25
24
20
21
20
24
22
24
23
23
19
21
21
22
20
22
51
51
45
40
41
45
41
45
35
40
37
40
51
57
43
44
57
60
41
40
25
48
49
26
44
48
35
31
41
63
46
44
53
54
44
40
39
46
43
43
37
45
36
40
50
60
44
44
NB: Public sector employees are defined as those working in the following sectors: public
administration, defence and compulsory social security; education, health and social work;
social and personal service activities (NACE Rev. 1 categories L, M, N and O).
Source: Rubery and Fagan 1998, Table 2.7, based on the European labour force survey, 1996.
Although the expansion of services employment (in particular the female dominated social
services) is a feature of all advanced societies, the pace of such expansion seems to be related
both to the type of welfare state, and to the stage of development of the economy. The nature
of the welfare state regime might influence women’s employment prospects directly in the
sense that a relatively large welfare state increases the demand for female labour. The reason
is closely related to the gender segregation of employment and occupational structure.
Overall, countries with a relatively low female employment rate tend to have relatively
underdeveloped social welfare services or low shares of private service work. The country
variation in the employment share in social services can serve as a first rough approximation
of the diversity of welfare state regimes. Hence, we may expect that the higher the share of
social services in the whole economy, the higher the female employment rates. As shown by
Figure 3.3 below, the correlation between female employment rate and the share of social
services (health, education and other social services) is high.
95
The job creation potential of the service sector in Europe
Figure 3.3. Female employment rates and employment share in social services (health, education and
other social services), EU-15 and the United States, 1997
Source: Eurostat (1999) and own calculations
Esping Andersen (1990, 1996) has suggested a country typology based on the extent to which
the market and or public sector provision has substituted for women’s unwaged domestic
labour in the provision of household-related services. The United States and the United
Kingdom are examples of economies in which the market delivers many services such as
ready-made meals and fast food outlets, laundry and home cleaning services, as well as some
childcare services and residential care for the elderly. Retail hours are relatively long, making
it possible to shop at the end of the working day. These private sector services are mainly
provided by women in low-wage jobs. In contrast, in the Nordic countries — and to a lesser
extent in France — many care activities have been collectivised and are carried out by women
employed in the public sector. Here, the public funding of care work has enabled women to
increase their labour supply, and it is precisely in the public sector delivery of this work where
large proportions of women have found employment. This has meant that the high female
employment rates in the Nordic countries coexist with high levels of gender segregation,
particularly between the private and public sector. Nevertheless, these public sector jobs are
generally better rewarded than the private sector equivalents found in the low-wage service
sector economies of the United Kingdom and the United States.
In the rest of continental western Europe, both public and private services are underdeveloped
by comparison. These include the Member States, such as Germany, with highly developed
welfare states, but where resources are distributed via transfer payments to ‘breadwinners’
combined with relatively generous derived rights for ‘dependent’ spouses, and the southern
European countries where welfare systems are relatively underdeveloped. In these countries
even more of the activities needed to run a home and look after family members are delivered
96
Service employment: A gender perspective
through women’s unwaged domestic labour than in either the Nordic or the Anglo-Saxon
economies. The impact on women’s employment rates can be illustrated by the example of
Germany, where the relatively low proportion of employment in the service sector is often
used to indicate that this sector is relatively underdeveloped. If Germany had the same sectoral
distribution as the EU average, the relative expansion of the service sector would mean that
female participation rates would increase on the basis of current female employment shares in
each sector (CERC 1994:95) (2).
Table 3.4 presents the female employment rates alongside the service sector employment
concentration rates for the Member States, grouped by typology.
Table 3.4. The female employment rate and the concentration of women’s employment in service
sector employment, 1996
% concentration of:
Public sector service model
Denmark
Finland
Sweden
Market service model
United Kingdom
Household service model
Belgium
Former East Germany
Former West Germany
Greece
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
EU-15
Women’s
employment
rate
Women’s
employment
in services
Women
Women’s
employees
employment
in the public sector in market services
67
58
68
83
80
86
54
50
60
29
30
26
62
85
44
41
46
56
55
39
32
52
43
36
35
55
59
54
50
85
80
78
62
80
82
80
72
92
88
76
65
80
53
44
40
43
39
46
43
43
37
45
36
40
44
32
36
38
19
41
36
37
29
55
43
40
25
36
NB: Total may vary slightly from the sum of full-time plus part-time due to rounding of decimal
places. See note to Table 3.2 for definition of the public sector services.
Source: European labour force survey, 1996.
(2) The counter argument is that the low proportion of employment concentrated in the service sector in Germany
simply reflects the fact that fewer activities have been outsourced from manufacturing. However, if we take the
proportion of men’s employment which is concentrated in service and sales occupations as an indicator of
service tasks rather than the sector they are employed in then we find that it is roughly three percentage points
lower than the EU average (Rubery and Fagan, 1998, Appendix Table 2.3a).
97
The job creation potential of the service sector in Europe
It is notable that the Nordic countries have high employment rates associated with a high
concentration of women’s employment in the public sector — what we have termed the public
sector service model — and the United Kingdom has a high female employment rate
associated with a high concentration of women in market services, what we have termed the
market service model. Female employment rates are generally lower in the other Member
States, but there is no simple linear correlation between female employment rates and the
relative size of either the public sector or market services.
While the market service model can provide a route to high female employment, the
expansion of public sector services has probably made a more positive contribution to
women’s labour market integration for two reasons. Firstly, the public sector has historically
been a relatively good employer compared to many parts of the private sector in terms of job
quality (security, pay, etc.), equal opportunities policies and the provision of more progressive
family friendly policies (such as parental leave, assistance with childcare services and
flexitime policies) (Rubery and Fagan, 1993, OECD 1994). Secondly, the public sector has
provided better quality services relative to those which most people can afford to purchase in
the market. Hence, women have an interest both as employees and as consumers in the
existence of regulated public standards.
However, common to both the market service model and the private sector model is the fact
that the segregated division of labour within the household around the traditional ‘male
breadwinner’ model of men as wage earners and women as homemakers is being replaced by
specialisation and segregation within the labour market. Women are over-represented in
services connected with providing physical and emotional care and personal services (health,
education, child and elder care, catering, cleaning, secretarial support, etc.). Gender
segregation in employment remains persistently high despite the feminisation of employment,
and women continue to do most of the domestic work within the home. From this perspective
the fundamental gender division of tasks within society have changed little despite the
substitution of state or market provision for some forms of domestic self-provisioning.
3.3. The implications of sex-segregated employment, service sector
restructuring and new work organisations for women’s labour market
integration
Women’s position in the labour market is dependent on the overall employment performance
of the economy, the employment structure and also the degree of sectoral and occupational
gender segregation. Ceteris paribus, the higher the level of gender segregation the higher the
gender impact of structural and sectoral change (3). Since the prevailing situation in most
advanced economies is characterised by a high level of gender segregation and that we cannot
expect large major changes in this respect in the short run, then further reduction of the gender
(3) For instance, the employment crisis in the early 1990s has been concentrated on manufacturing industry and
construction, both sectors where the gender distribution of employment is very uneven (male-dominated
sector).
98
Service employment: A gender perspective
employment gap will be essentially related to the growth of female-dominated sectors. In
other words, the further integration of females in the labour market, given the level of
segregation, is highly dependent on the employment growth in female-dominated activities.
As stressed by Fagan & Rubery (1993), the patterns of occupational and sectoral segregation
have in recent years tended to provide some protection for women’s employment, both
because women have been concentrated in areas with significant job growth and because these
sectors have been less prone to cyclical variation. At the same time, occupational segregation
is a key factor in the persistence of gender inequality within the labour market. Segregation
facilitates the maintenance of a gendered labour market in which female-dominated and maledominated jobs are organised according to different principles including pay, working time,
and integration into career ladders and organisational structure (Rubery et al., 1999).
Past trends in service sector expansion suggest that further expansion of public and or market
services would help to raise the female employment rate in countries where it is currently low.
This is because European employment growth in recent decades has been concentrated in
those service activities where women already had a reasonable presence, and this has played
an important role in contributing to women holding a growing proportion of jobs in Europe.
Indeed shift-share analysis shows that the feminisation of employment in the period covering
the 1980s and early 1990s was due in roughly equal parts to the expansion of employment in
female-dominated sectors (mainly services) on one hand and an increased female share of
employment in male-dominated sectors on the other. While the latter process indicates some
reduction in gender segregation, the increased female share in female-dominated sectors
means that there has not been any overall dismantling of gender segregation (Rubery et al.,
1998, 1999).
The gender segregated basis of women’s gains in employment is also evident in occupational
trends associated with economic restructuring. Women have benefited from the expansion of
professional employment in recent decades, much of which is in the service sector, and hold
a higher and growing proportion of these high-status and high-paid occupations in the
economy. Some of these professional jobs were previously male-dominated, such as
accountancy or law, while others were already mixed or female-dominated, such as teaching
and nursing. This feminisation of the professional tier of the labour market can be considered
as a process of desegregation, even if other sex divisions persist or are emerging to maintain
vertical segregation within the senior grades and managerial positions. However, segregation
has increased at the intermediate and lower levels of the occupational hierarchy. Femaledominated occupations such as clerical work or personal services have been further feminised
as women have increased their share of employment in these parts of the economy. Few
women have entered male-dominated manual jobs, not least because job loss has been
concentrated on this part of the economy since the 1970s so that there have been few openings
or incentives for women to compete in this area. Summary index measures present a picture
of persistently high overall levels of segregation in each Member State, with women in parttime jobs even more segregated from men than women in full-time jobs (Rubery and Fagan,
1993, Rubery et al., 1999).
99
The job creation potential of the service sector in Europe
While service sector expansion has enabled women to increase their share of employment on
a gender segregated basis in the past, current labour market developments suggest that the
gains women have made may be stalled or reversed. These labour market developments can
be summarised as follows, although the extent and form of these general trends varies by
sector and society:
• the deregulation of employment protection and increased flexibility in contractual and
working time arrangements;
• the decentralisation of collective bargaining, a decline in union representation and increased
individualisation of the employment relationship;
• organisational restructuring to take advantage of new forms of technology and work
practices is producing organisations that are characterised by flatter hierarchies, greater
functional flexibility, a ‘hollowing out’ of intermediate-level jobs, and the collapse of
internal labour markets with rigid promotion ladders in favour of career progression based
on qualifications and lifelong training;
• persistent unemployment among those with few qualifications.
To start with, a potential development is that of more direct competition with men for jobs than
in the past. While professional employment is still expanding overall in most countries,
organisational downsizing and delayering is producing job loss and reduced promotion
opportunities in the middle management levels at the workplace (Capelli et al., 1997, Pryor
and Schaffer, 1999). This may reduce a number of routes which women have used to move
into management in the past. In addition, the link between qualification levels and career
progression may be becoming less certain in the new flexible workplaces of constant
innovation to survive in changing product markets. This may have negative implications for
women, because so far they have made more progress into the higher status and better paid
job areas where qualifications and examinations have played a large role in selection, such as
in the professions and the public sector in many countries. Where selection criteria are less
transparent, notably in private sector managerial jobs, women have made less progress
(Rubery et al., 1999). Furthermore, as labour markets restructure and entry qualifications
become more rapidly out of date, access to workplace training and lifelong learning will take
on more importance in some labour markets. Yet women’s ability to access workplace training
and lifelong learning may be constrained by segregated employment patterns and career
interruptions connected with child raising, for lifelong learning opportunities tend to be
offered primarily to those in full-time, continuous employment (Tuijnman and Schömann,
1996).
More intense competition might occur among those with few qualifications because the
shrinkage of manufacturing has reduced the traditional job opportunities for men. In the longterm, women can expect more men to try and enter their traditional enclaves in clerical and
service work, particularly where reform in unemployment benefits and active labour market
policies are putting more pressure on men to accept jobs in these areas. The experience of East
Germany following unification provides a clear, albeit, extreme warning. Here, the
combination of high unemployment and the rising status of clerical work under the new
100
Service employment: A gender perspective
regime led men to take refuge in what were traditionally women’s areas of work (Maier et al.,
1996:35). Furthermore, low-qualified women may find themselves increasingly in
competition with other pools of labour for part-time work in retail, hotels and catering. For
example, university education in the United Kingdom has expanded in the context of the
removal of student maintenance grants, so that now approximately one fifth of students hold
part-time jobs. Another substitute labour pool may be among those who have retired from
full-time work but seek part-time employment to supplement their income or for the nonfinancial rewards of work.
Employment prospects will continue to be affected by developments in information and
communications technology (ICT) which play a vital role in the growth and organisation of
service employment. This technology could be seen as the first major technological
innovation that has directly addressed parts of the service sector, given that many service
activities have essentially to do with the processing of information. This new ‘informational
paradigm’ (Castells, 1996) is transforming the organisation of work and producing new
services. Technological change and organisational restructuring appears to be reducing
employment opportunities in clerical work, which until now has provided an important source
of middle-level jobs for women. Over the 1990s some countries recorded downturns in the
rate of growth of this occupational group or even actual declines (Rubery et al., 1999). Part of
the reason is that clerical functions are increasingly being carried out by workers in other
occupations using personal computers and other information technology, thus enhancing their
functional flexibility. Another important reason is that large proportions of clerical workers
are employed in financial services, where jobs are being cut back in a number of countries as
a result of organisational rationalisation and mergers, and the introduction of new forms of
service delivery such as telephone and Internet banking.
At the same time that information and communications technologies are producing job loss in
some areas, new services and areas of employment are emerging. Women can be expected to
gain employment in these new service jobs where there are low or intermediate level jobs,
such as in telephone ‘call’ centres, but their prospects in the more highly skilled areas are less
clear. For example, the emergence of computer professionals as a new and expanding
occupational category in recent decades has gone hand in hand with this becoming a maledominated area of work, although women are making some small inroads in some countries
(Rubery et al., 1999). The potential for expanding telework is also double-edged. On one hand
it offers the scope for more flexible and possibly more family-friendly ways of organising
working life, as well as spreading employment into economically depressed or isolated
geographical areas. On the other hand, it carries the risk of placing an increasing number of
women at home working in isolation and with inferior terms and conditions of employment if
this form of work is used to drive down labour costs (Huws, 1995).
The quality of women’s jobs in public and private services may also deteriorate in the context
of current labour market developments. Where public sector employment has provided
women with reasonable employment conditions in the past this may be undermined through
restructuring processes in the context of expenditure constraints driven by fiscal convergence
criteria. These budget limits are constraining public sector employment growth and also
101
The job creation potential of the service sector in Europe
fuelling the restructuring of service delivery. In some circumstances the direction of change is
to try and ‘roll back’ the welfare state with the aim that the market will expand to provide
substitute provision, for example, in elder care services and private pension schemes. In other
circumstances new accounting systems have been introduced to encourage increased
competition between service providers within the public sector and between the public and
private sectors through processes of privatisation and competitive tendering. The
subcontracting of services from the public to the private sector in the United Kingdom, for
example, was accompanied by deterioration in employment conditions for employees in these
services, most of whom are women. The application of the transfer of undertakings’ protection
of employment directive to public sector workers has provided some protection by requiring
the subcontractors to respect minimum terms and conditions of employment (Escott and
Whitfield, 1995). However, in the long run the real value of these minimum guarantees may
fall, not least because public sector conditions have themselves been pegged to limit the
growth of the wage bill. Women’s dependence on future employment trends in the public
sector may be particularly marked in labour markets where there is little mobility between the
private and public sectors. This has been predicted in the case of Denmark, for example (Boje,
1995:33).
Developments in the organisation of private services may also have negative implications for
the quality of women’s employment. Women’s private service sector employment is
concentrated in the least regulated parts of the economy, such as in small firms. Many of them
are found in the lowest paid, highly feminised areas of private service activity, and patterns of
gender segregation are even more pronounced when viewed from the establishment or
workplace level than from the aggregate sector or occupational level (Burchell, 1996,
Millward and Woodland, 1995, Scott, 1994). This makes women vulnerable to deterioration
in wage and other employment conditions in the context of current trends towards
deregulation and decentralised bargaining.
Finally, the expansion of part-time and other flexible employment contracts in services has
been associated with the mobilisation of mothers into employment in some Member States.
However, the creation of such jobs is not necessarily a requirement for raising the female
employment rate. This is shown by the fact that in some Member States large proportions of
mothers are employed on a full-time basis (see Figure 3.1 above). The quality of part-time
work varies markedly between countries and sectors in terms of job type, employment
conditions and promotion opportunities. Where part-time work has developed into major
employment within a strongly regulated labour market there are fewer penalties associated
with working part-time in terms of pay, access to benefits and job opportunities further up the
occupational ladder. Examples of these relatively positive findings include the Nordic
countries and the Netherlands (Fagan et al., 1995, O’Reilly and Fagan, 1998). Nevertheless,
large proportions of part-time jobs are concentrated in private sector services which generally
have inferior pay and employment conditions, often organised around very short or marginal
hours contracts. Further expansion of this form of employment will increase the polarisation
between women in full-time and part-time jobs, particularly where there are few mobility
channels to move between part-time and full-time jobs in either the external or internal labour
market.
102
Service employment: A gender perspective
To conclude this section, future employment growth is predicted to be concentrated in service
activities. Many of these jobs will be connected with welfare services (health, social services
and education) as well as personal and household related services. How this expansion in
welfare services will be funded, and whether it will be organised through direct public sector
provision or through combinations of public, private and voluntary sector partnerships is not
clear. If the outcome is a reduction in public service provision driven by public expenditure
constraints to meet the convergence criteria of monetary union, then women will be affected
not only through a potential deterioration in both quantity and quality of public sector
employment but also because if the public provision of childcare and elder care is cut back,
this will constrain their labour supply. The quality of many private sector service jobs are also
under threat due to deregulation and organisational restructuring. In addition, competition is
likely to increase between women and men in many parts of the labour market. Those with
few qualifications will continue to bear the brunt of unemployment and may face even more
limited job opportunities if competition within the labour market produces a ‘bumping down’
process of workers being forced to compete for jobs below their qualification levels (Pryor
and Schaffer, 1999). While the expansion of part-time and low wage service jobs has been
posited in policy debates as one means of job creation to relieve unemployment, what is often
missing is the recognition that only people with access to other household resources can afford
to take such jobs. Typically, these are women with employed partners, young people, or retired
persons with a pension. Such forms of work are unlikely to be voluntarily entered by
unemployed men with family responsibilities, undermining the efficacy of promoting parttime work as a means of tackling unemployment (Walwei,1998). To the extent that women fill
these jobs the effect will be to maintain the sex-segregated character of the labour market in
which employed women are disproportionately concentrated in the low paid parts of the
economy.
3.4. The gender division of labour, the service economy and equal
opportunities: A micro perspective
A growing proportion of women are remaining in the labour market throughout their working
lives and combining waged work with raising children. From the perspective of household
time allocation, this means that women are devoting more of their time to the activity of
employment, leaving less time for domestic production and leisure. Yet in the gender division
of household labour women still do most of the day-to-day housework and childcare. This
domestic responsibility constrains women’s labour supply, and the type of jobs women take
are at least partly influenced by the working time requirements. As discussed in the previous
section, the substitution of market and public household related services, particularly that of
childcare provision, has facilitated the increase in women’s employment in recent decades. In
this section we consider the relative efficiency and the gender implications of market subsidies
to stimulate expansion of household related services. The main objective is to explore, at a
theoretical level, to what extent such a policy may improve gender equal opportunity, overall
economic efficiency, employment prospects and economic growth.
103
The job creation potential of the service sector in Europe
3.4.1. Gender and time allocation in the household
In the household economics approach to time allocation (Becker, 1965), the household is
viewed as a ‘small factory’ in which the labour of household members is combined with the
purchase of market goods and services in order to nurture and sustain the household. Time is
used in consumption to varying degrees of intensity. Some forms of consumption are time
intensive, in that they require a substantial input of domestic labour to transform the purchased
inputs into the final goods and services (for example, shopping for and cooking a meal,
looking after children, home decorating, making clothes). Other ‘labour saving’ forms of
consumption require less time to be allocated to domestic labour (for example, ready-cooked
meals, childcare services, hiring a painter), in other words market purchases are substituted
for domestic production. The pattern of market consumption which results is therefore subject
to the constraint of income for market purchases and available time for domestic labour and
consumption.
The implications of this model of time allocation are that if earnings rise then the financial
rewards from allocating time to employment increase and conversely allocating time to
domestic work and leisure becomes more expensive in terms of foregone earnings. As a result,
labour supply will tend to increase, ceteris paribus. Conversely, if the prices of certain market
goods and services fall then there will be a tendency for these to be substituted for domestic
production, thus freeing time which may either be reallocated to other preferred forms of
domestic activity (leisure instead of housework) or may feed into an increased labour supply.
Given the household budget constraints of time and money it is assumed that the gender
division of labour found in couple-households reflects an efficient specialisation to capitalise
on comparative advantages in waged and domestic work. The assumption is that it is more
efficient for the lower earning member (typically the woman) to specialise in domestic
production and for her higher paid partner to devote more time to waged work.
The value of the household economics approach is that it emphasises the trade-off of time
allocation between waged work, domestic production and leisure. However, the claim that a
gender-based division of labour within households is efficient is erroneous. A number of
market failures are exposed when the returns to the household are viewed from a lifetime
instead of cross-sectional perspective, and when the aggregate macroeconomic gains of
increased female integration into the economy are taken into account (Fagan and Rubery,
1996). In brief, policy intervention to enable women to increase their labour supply would
enable women’s skills to be more fully developed and used in the economy, thus increasing
economic efficiency. This enhanced integration of women into the economy would also
provide income security in periods of male unemployment and household dissolution due to
divorce or widowhood, thus reducing the demands on public expenditure for unemployment
and poverty assistance. This suggests that if market subsidies of household related services
promote a shift towards a ‘dual-earner’ model of household behaviour then such policy
intervention may enhance equal opportunities and economic efficiency. This idea is explored
further in the rest of this section of the chapter.
104
Service employment: A gender perspective
3.4.2. The gender division of domestic labour, market substitutes and equal
opportunities
Women still do most of the unwaged work in the household despite their rising labour market
involvement. Some research has revealed a lagged adaptation whereby men are gradually
increasing the amount of unwaged work they do in circumstances where the woman works
full-time, with the evidence being more pronounced in households where the women and men
possess high qualification levels (Gershuny et al., 1994, Vogler, 1994, Rubery et al., 1998,
Fagan, 1997). Adaptation follows rather than precedes women’s integration into waged work,
very slowly and often lagging across generations (Gershuny et al., 1994).
How then might subsidised market services affect the gender division of domestic labour? If
the price of household related market services fell as a result of introducing subsidies then this
could be expected to produce a substitution of market consumption of services for domestic
production. This would enable women to allocate more of their time to other activities. This
might take the form of increased leisure or family time in some societal contexts, and in others
women might devote more time to employment. The largest impact of market subsidies on
increases in women’s labour supply could be expected in countries where the female
employment rate is low, or largely associated with part-time work.
However, even where subsidised services did not lead to a direct increase in the volume of
women’s labour supply it may have three positive outcomes by reducing the time pressures
that women are under when undertaking employment in a gender divided society. Firstly, it
may make it easier for women to combine employment with domestic responsibilities, thereby
enabling them to hold onto their occupational position and to advance up career ladders. This
would produce a higher utilisation of women’s skills in the economy. Secondly, reduced time
pressures will reduce stress and other negative health effects associated with having too much
work to do, and will enhance the quality of family life, producing gains for children,
enhancing marital relationships and so forth. Thirdly, consumption of time-intensive leisure
activities may increase if households had more time, contributing to job creation in certain
parts of the service economy. Some research does show that when people have more time (and
energy) they have more active leisure activities (Olsson, 1999). For example, they are more
likely to take up evening classes and other forms of lifelong learning, sport and attending
cultural events.
Thus one benefit of an expansion of household-related market services may be to promote
equal opportunities by allowing women to reduce their domestic workload by purchasing
market substitutes and so enable them to compete with men on more equal terms in the labour
market. Another related benefit is that the female employment rate could be expected to rise,
for a large proportion of the workforce in these household-related market services is likely to
be women on the basis of existing gender segregation in the labour market. However, the
quality of the jobs which emerge will also have an impact on the situation of women in the
labour market. If the jobs which result are low paid or precarious in other ways then this will
contribute to the developing polarisation between highly qualified, well-paid women and
those who are less qualified (Rubery et al., 1999). In effect, the result will be to create
105
The job creation potential of the service sector in Europe
subsidised, low-waged work for women with few qualifications, who will provide services to
households containing well-paid women.
Even if the expansion of market provision of household-related services is organised around
better quality jobs, such as emerged within the public sector model of service provision, this
will do little to challenge the existing division of household labour between women and men,
given the slow adaptation of men to involvement in domestic work. To promote equal
opportunities, other complementary policies are needed alongside subsidised household
related services, targeted at increasing men’s involvement in domestic labour. Policies which
encourage men to become more involved in domestic work include working time reductions,
parental leave entitlements which are targeted at fathers, and more general educational
campaigns to shift social norms (Fagan and Rubery, 1996).
3.4.3. The structure of wages, taxes and benefits and the effect on the consumption of
market services
The size and organisation of market services in an economy is partly dependent on the wage
structure and earnings distribution in the society. For example, countries with a narrow wage
dispersion and high average income can be expected to produce a rather different form of
consumption of market services than one with high wage inequality (Esping Andersen, 1996).
In the Nordic countries, for example, the labour market is characterised by a high female
participation rate, a large proportion of which is in the public sector, some employment in
household-related private sector services, a low wage dispersion and a relatively high average
income. It has been argued that the combination of high average income, a low earnings
dispersion, and high taxes precludes the development of private household-related services.
According to this view, the high tax wedge constitutes a clear hindrance to the development
of such market services in the formal economy, and encourages the expansion of the informal
economy.
In contrast, the wide dispersion of wage levels found in the United Kingdom and the United
States means that the relatively high disposable incomes of the professional classes makes the
market purchase of household-related services more affordable for this section of society. In
effect, it means that the low-waged work of women with few qualifications provides the
means for professional couples to ‘buy-out’ of renegotiating the gender division of household
labour between themselves and to substitute the labour of women from other households.
The gender division of time allocation is also affected by the prevailing tax and benefit system.
An important issue related to the relative deficit of employment growth between Europe and
the United States has been the greater attention paid to the functioning and the incentive
structure of the labour market. Social transfers, taxes and non-labour costs may, either
separately or jointly, raise barriers and disincentives affecting the growth of service activities.
In particular, a high tax wedge has been found to restrict consumer demand for householdrelated services, which limits job creation in service activities and leads to inefficiency in
resource allocation within the household and the aggregate economy (Henrekson, 1998,
106
Service employment: A gender perspective
Pålsson, 1997). These wedges may affect household decisions in terms of the gender
allocation of time and labour supply (hours worked), giving rise to economic inefficiency.
Econometric studies show that reform of income taxation could contribute to increasing the
labour supply of some groups of women, as labour supply is relatively more responsive for
married women at the point of entry to the labour market and for those in short part-time jobs
(OECD 1994). Yet national differences in married women’s participation patterns are only
loosely correlated with effective tax rates (Vermeulen et al., 1994, CERC, 1994). This
suggests that the impact of joint taxation on women’s labour supply decisions is likely to be
weaker than the constraints of childcare or the high effective marginal tax rates which arise as
means-tested benefits are withdrawn.
3.4.4. Women’s employment and the consumption of market services
The increased involvement of women in waged work may itself generate job creation through
higher levels of consumption to substitute for women’s unwaged labour (ready-prepared food,
laundry services, childcare services) and more generally due to the high living standards that
dual-earner households have compared to most sole-earner households. The extent to which
this market substitution occurs will be influenced by relative income and price levels, as well
as social conventions concerning appropriate consumption patterns.
Once total household income is controlled, the expenditure patterns of dual-earner and soleearner households are broadly similar in the United Kingdom, suggesting that there is no
simple process of market substitution for self-provisioning (Horrell, 1991). Dual-earner
households do purchase more prepared food (Horrell, 1991), and they are also more likely to
purchase certain services, notably childcare and to a lesser extent home cleaning and
maintenance (Gregson and Lowe, 1994, Warde, 1990). Similarly in Germany there is a slight
increase in the amount of household services purchased in two-earner or high-income oneearner households, accompanied by a rapid increase in the number of marginal part-timers
employed in such services, particularly in private households. In Germany this consumption
is likely to be in addition to self-servicing, rather than as a substitute, for the volume of unpaid
work in households has remained stable (Maier and Rapp, 1995:89). However, even if there
is no direct difference in consumption patterns between dual earner and other households, it
is likely that over time women’s participation has affected consumption patterns generally, so
that changing societal norms, rather than simply whether the woman is employed or not,
influence household expenditure patterns on household appliances, convenience foods, etc.
3.4.5. The job creation possibilities of expanded market provision of household-related
services
Even if public welfare and the consumption of market services have substituted for some part
of domestic production within the household, unpaid domestic work continues to make a
substantial contribution to economic production. It is, for example, in Germany, estimated to
amount to 38 % of gross national product (GNP) (Maier and Rapp,1995:86) and to use a larger
volume of labour than that used in the formal market economy of the Netherlands (Plantenga
107
The job creation potential of the service sector in Europe
and Sloep, 1995:71). The evidence from time budget studies suggests that there is only limited
scope for reductions in domestic labour when time allocated to employment is increased,
particularly when there are children to look after. There is some reduction in time spent on
housework, probably due to fewer meals being eaten at home (Horrell, 1994) as well as a drop
in self-imposed standards (Oakley, 1974). Overall, the main difference is that women have
less time for leisure, sleep and to spend with their children (Horrell, 1994:212); in other words
they take on the ‘second shift’ (Hochschild, 1990).
Given the vast amount of productive work that is done in the domestic sphere, what are the
job creation possibilities of expanded market provision of household related services? A 1990
Dutch study explored the unfulfilled service needs of dual-earner households compared to
sole earners, on the basis that the first category has more income but less time (study cited in
Plantenga and Sloep, 1995:72-3). The study found that the main concern of dual-earner
households was for more convenient store opening hours, although few were prepared to pay
more for extended services. Another main demand was for domestic help, but people were
generally unwilling to pay more than the usual informal rate. A quarter of those with children
had unsatisfied childcare demands, mainly due to long waiting lists for existing centres. Even
on cautious estimates the study showed that there was a demand in principle for some
domestic activities to be outsourced, but the problem was that the prices households were
willing to pay were lower than market rates within the formal economy. The authors conclude
that either the quality of formal services have to be raised so that consumers are willing to pay
more, or prices have to be reduced through subsidies or reductions in labour-related costs. One
suggestion was that a portion of social security benefits could be retained as an in-work benefit
to subsidise certain activities — such as childcare and home cleaning — which are not in
direct competition with existing formal services.
This study of informal work in the Netherlands suggests that the inability or unwillingness of
many households to pay full market rates for some services is a constraint on the development
of these activities in the formal economy. Trading of these activities may increase in some
countries in the informal sector or in the formal economy where private services can be
delivered relatively cheaply. The latter is likely to occur in a United States and/or the United
Kingdom type scenario of highly dispersed income levels, where private services are
delivered by low-paid employees and consumed by the high-income professional households.
A growing number of initiatives to expand the market for domestic services have been
launched or proposed in the Member States. These policies are designed to enhance
employment growth in household-related services, and can be classified into two broad
categories:
• supply-oriented policies aiming to encourage the creation of enterprises in the service sector
(start-up schemes etc.),
• demand-oriented policies aimed at increased household demand for service activities (price
and fiscal subsidies etc.).
108
Service employment: A gender perspective
To illustrate, a 10-year policy experience exists in France consisting of a tax exemption
(chèque service) for the purchase of domestic services. Evaluations of initiatives to expand
the market for domestic services (mainly from Belgium, Denmark and France) show that the
net macroeconomic impact on employment and unemployment is low. The main effect is to
‘formalise’ services through shifting informal economic activity into the formal economy
(Anxo & Tanemar, 1996, Cette et. al., 1998). However, such evaluations generally fail to
address the impact of such services on the gender allocation of time in households. It is
important that gender issues are mainstreamed into future evaluations of such initiatives.
Furthermore, from a longer term gender perspective, it may be necessary for policy initiatives
to stimulate market provision to substitute for a decline in family and informal service
provision. There are already indications that some areas of informal and voluntary care work
currently carried out by women may decline due to their changing employment patterns. More
generally, the processes of urbanisation, individualisation and geographical mobility in
modern societies (Beck, 1992) are changing the types of responsibilities and expectations that
family members and neighbours have for each other.
3.5. Policy implications
National differences in the structure of employment, in particular the relatively large country
variation in the relative size of the public and market services and the large variation in
women’s employment rates have several policy implications. First, this diversity means that
there are differences in the potential for female employment growth in traditional female
employment areas between Member States. To illustrate, the high concentration of women’s
employment in social services in the northern countries seems to be reaching saturation point,
while in some southern countries, this share is still largely below the European average and
thereby leaves scope for public intervention. Hence, public policies promoting the
development of employment in social services (health, education) and the outsourcing of
female domestic tasks into the market (childcare in particular) may accelerate women’s labour
market integration by transferring some tasks from the domestic sphere to the labour market.
This development would have a positive impact on the overall employment rate and reduce
the prevailing gap between the European Union and the United States. In particular, the tight
public expenditure constraints as part of the economic and monetary union (EMU) is
constraining the scope for countries with underdeveloped welfare systems to ‘catch-up’ with
other European countries. Consideration should be given to providing scope for public
expenditure expansion on service provision in these countries. Generally, consideration
should be given to ways of encouraging the restructuring of existing public expenditure away
from transfers to ‘breadwinners’ to support their ‘dependent’ spouses and into spending on
service provision, to improve the long-term employment and fiscal base.
The second policy implication is that any expansion of services that are heavily dependent on
female labour will draw more women into employment but do little to reduce the sexsegregated structure of employment. As Rubery et al. (1999) argue:
109
The job creation potential of the service sector in Europe
‘This dual role of occupational segregation in promoting the demand for female employment
but at the same time maintaining and reinforcing gender differences, presents a major dilemma
for policy toward equality’.
A two-pronged policy to address segregation is required. Firstly, anti-discrimination and
positive action measures are required to enable women to move into the better quality areas
of male-dominated employment from which they have been excluded. Secondly, the quality
of the jobs which women currently do also needs to be improved. It is also important to
recognise that the lower status of women within the occupational hierarchy is partly a
reflection of the undervaluation of women’s work, as well as the failure to create job ladders
and career structures within female-dominated areas. One aspect of this is regulation to protect
basic employment conditions (job security, minimum wage etc.) to prevent a deterioration in
employment conditions for women (and men), thus creating a safety net for those in the most
vulnerable labour market positions. The other aspect is that positive action should not simply
involve measures to move women into male-dominated areas. It should also involve a
reconsideration of the organisation of the job hierarchy to provide greater recognition of the
skills involved in many female-dominated job areas, and to create links from these job areas
into mainstream promotion lines. The current context of trends towards flatter hierarchies and
more functional flexibility in many organisations may provide more scope for such tactics of
re-evaluation than in the past. Indeed, on a more negative note, these organisational trends
may make such tactics critically important, for otherwise women’s jobs may be enlarged to
encompass more responsibilities but without the enhanced pay, status or access to promotion
channels commensurate with the changes to job content.
The third policy implication is that the quality of part-time work in the service sector needs to
be enhanced, both to improve conditions for women employed in such jobs and to make these
jobs more attractive to unemployed men. As well as ensuring compliance with the principle
of equal treatment, fiscal policy and labour law should be reformed to remove exemptions for
jobs with low earnings or few hours. These exemptions provide subsidies and incentives for
employers to create short-hour part-time jobs rather than longer part-time jobs, which are more
at risk of being marginalised forms of employment. It is when part-time jobs are organised
around longer hours that they are more likely to be integrated on an equal basis alongside fulltime jobs at the workplace and in the social protection systems (O’Reilly and Fagan, 1998).
The fourth issue is the under-representation of women among the self-employed. The share of
women engaged as family workers is declining as their involvement in waged work rises, but
there is less evidence of a trend towards equal representation in self-employment. On this
basis, action is needed to boost women’s involvement in entrepreneurship, particularly in the
southern countries where such activities are a major part of economic life. Furthermore,
women’s self-employment may be preferable to low-wage service jobs for large companies
such as retail conglomerates or other large firms, which provide household-related services
such as contract cleaning. Enabling women to set up businesses may mean that they develop
businesses to deliver childcare and other household related services. This may contribute to
the expansion of such market services as well as providing women with better opportunities
for developing their own autonomous careers.
110
Service employment: A gender perspective
The fifth issue is that the quality of services provided, whether by the public sector or by the
market, must be guaranteed, particularly in relation to child and elder care. If the services are
perceived to be of low or deteriorating quality then households will be reluctant to use them,
particularly where there are strong cultural norms that such activities should be done by
women within their homes rather than outsourced. Where high quality public sector provision
exists it is important that any expansion of market services does not lead to a reduction in
public sector provision or a deterioration in public sector standards.
111
Chapter 4
Labour costs, social habits and employment
regimes. Two comparisons of national
employment performances and some lessons
Françoise Benhamou and Bernard Gazier
4.1. Introduction
The available literature on the comparative analysis of national employment performance can
be roughly organised along two dimensions. The first is the starting point: one may start from
a sizeable gap in employment performance and try to explain it; or one may start from more
or less equivalent employment performances at a given time and deepen the underlying
dynamics. The second is the kind of causality put into the forefront. An ongoing debate
persists between the economists who insist on labour costs (either in crude, or sophisticated
and indirect ways) and those who put the emphasis on non-cost determinants such as societal
habits.
The main orientation of this contribution is to start with such divisions and to show that other
determinants and a causality network can usefully be introduced in order to better the
understanding of different employment performances and of policy perspectives.
The organisational and institutional context surrounding the work relationships and the risk
management arrangements which are behind them appear to be important elements to
consider (1). They may help to interpret the facts as well as to assess policy orientations. In
particular, we aim to show the relevancy of the hypothesis of different national flexibility
regimes (namely, employment arrangements and behaviour securing flexibility) which rely on
different tools and institutions.
(1) Cf. Gazier and Thévenot in this volume.
113
The job creation potential of the service sector in Europe
We selected two comparisons:
— retail trade between France and the United States, which exhibits a considerable
employment gap,
— cultural employment between France and the United Kingdom, where the quantitative
differences are not important but where the underlying dynamics are.
The choice of these countries and activities is of course partly dependent on the availability
of data. However, if one chooses to focus on employment regimes, it is interesting to examine
countries with strong societal differences as regards the treatment of the work force and social
protection. In our cases, one may oppose the French welfare state regime, which has been
termed corporatist-continental, and the United States and/or the United Kingdom welfare state
regimes, which have been termed liberal welfare states. Our comparisons allow us to consider
national situations and labour markets characterised by very different institutional settings,
and to examine some possible employment implications.
In the first case (France/United States, retail trade), a well-developed controversy exists
regarding the enormous French gap (or lag) compared to the United States employment level.
Intensive efforts have been made to explain the situation, and it is useful to take a closer look
at the arguments presented by each side in order to introduce the risk management and
institutional context perspectives (Section 1). As regards cultural employment, the starting
point is the reverse; the employment levels do not seem that different in France and the United
Kingdom. However, important qualitative differences are to be taken into account to obtain
some understanding of the situation and the prospects. It will be shown more extensively by
gathering and interpreting various data, and assessing a number of different hypotheses on the
determinants of job growth in this sector (Section 2).
4.2. Employment in the retail trade sector in France and the United States:
An appraisal of the Gadrey/Piketty controversy
4.2.1. Setting the problem: Evaluating the potential job creation of France in the retail
trade sector
In 1996, there were 127 million jobs in the United States of America, and 22 million in France.
The United States population was 4.5 times greater than the French population. The
employment rate (number of jobs per person) was 25 % higher in the United States. It
corresponds, for France, to a gap of 5 million jobs (such a comparison is not fully valid of
course, but gives a kind of maximum benchmark). T. Piketty (Piketty, 1998, p. 81) projects the
United States job structure and employment rate on the French case. The figures are: Retail
trade, actual French employment 1996: 2 million; according to the United States structure: 3.7
million. The same gap holds for the hotel and restaurant sector: actual French employment
1996: 0.77 million; according to the United States structure: 1.7 million.
114
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
These differences account for more than half of the ‘job gap’ between France and the United
States. Of course, numerous complicated processes must be introduced in such a ‘catch-up’
process, as, for example, the rise in activity rates induced by rapid job creation. One job
created is not equivalent to the transformation of an unemployed person into an employed
person. Nonetheless, the comparison is striking. Even if there is room for discussion on the
job yardstick, it is clear that these two sectors are more developed in the United States than in
France, and there is scope for job development.
According to these evaluations, in the retail trade, the employment level per person is actually
80 % higher in the United States than in France.
J. Gadrey and J. Jany-Catrice agree, and present a slightly different evaluation of the same job
gap for retail trade (Gadrey and Jany-Catrice, 1998 p.102). They remove the automobile retail
trade, and compute the employment rate by introducing the total number of persons in each
country. Then the gap is 68 %, which was finally accepted by T. Piketty.
There is another important agreement; a significant part of the gap stems from differences in
the consumption levels. The American people consume more products, and the volume of
purchases is 30 % higher in the United States. So, following Gadrey and Jany-Catrice, the gap
which remains to explain is 29 % (1.68/1.30). The United States retail trade firms use 29 %
more work than in France, for selling the same product volumes. T. Piketty agrees (although
he presents some critical remarks) with this evaluation.
A last field of quasi-agreement is the differences of wage costs between France and the United
States in the retail trade. Thomas Piketty 1997 set an evaluation of 40 % more in France, while
Gadrey and Jany-Catrice proposed 20 % more. In his last article of 1998, Thomas Piketty
agreed more or less with this second evaluation, and introduced it together with the previous
one while suggesting that more detailed investigations would be necessary for reaching a
definitive agreement.
4.2.2. The divergences
Thomas Piketty identifies a major candidate for explaining the 29 % difference; the wage costs
in a sector particularly intensive in low-skilled work. He argues that a 20– 40 % difference in
wage costs is enough to explain most of the differences of job performance between the two
countries.
In the absence of a formal demonstration, he presents two series of arguments in favour of his
thesis. The first are rooted in a very long-term perspective. He observes that most of the
relative employment level in trade services was obtained in the USA before the 1960s (share
of jobs in the trade sector among all jobs — 17.7 % in 1960, and 17.5 % in 1996), while the
French evolution, catching up since the beginning of the century, lagged since the 1970s
(10.3 % in 1954, 13.2 % in 1973 and 13.6 % in 1996). He suggests that the lag was created
since the 1970s for reasons of rising labour costs. Second, he presents case study evidence that
the costs of labour are considered by employers to be an obstacle to more employment in the
115
The job creation potential of the service sector in Europe
sector. The consequence, in his view, is that a policy of cutting social security taxes for lowwage workers could massively improve the job performance in this sector, without harmful
social side effects.
In contrast, J. Gadrey and F. Jany-Catrice consider that the wage cost explanation is not valid.
They begin by introducing working time. The working time per person in the retail trade is
different between the two countries, with more part-time jobs and fewer full-time workers in
the United States. They estimate that the effect of this factor accounts for 6 % of the extra work
volume in the United States. So, the problem becomes explaining a 22 % difference
(1.29/1.06). The United States retail trade uses 22 % more work than retail trade in France for
the same product basket.
They reject the premise that part of the explanation could be sought in productivity levels;
taking into account the numerous difficulties for evaluating service quality, the United States
level could be even slightly higher (5 %) than the French level. The difference corresponds
to additional labour-intensive selling services given to the United States consumers;
home delivery, packing assistants, doorkeepers, more opening hours, counselling and
information, etc.
These additional services are not financed by the consumers, who pay more or less the same
price for the same product basket in both countries, but by the workers themselves. Here, we
find the labour costs put in quite a different perspective. The United States retail trade workers
are twice as inexpensive for their employers. Compared with the French situation, in
purchasing power parity the United States workers’ mean annual gross wage is 14 % lower,
and the lower employer national insurance contribution level corresponds to an additional
lowering of total labour cost by 11 %. So, the United States employees receive lower wages,
and also pay individually, those expenditures (illness, retirement) that are collectively
financed and part of the wage bill in France. They finance twice the production and the sales
of extra services in retail trade. The point is that such a cost structure depends on a collective,
organisational and cultural agreement — at the expense of the United States workers.
As a result, Gadrey and Jany-Catrice argue that a tax cutting policy cannot have a big impact
on the job volume content of retail trade activities in France. First, they observe that during
the 1990s, important tax cutting was already realised in France. This centred on low wages,
which mainly benefited the trade sector being the most important low-skilled labour user, and
having no apparent effect on the job volume performance. The main reason for this is that the
competition structure of the sector as well as the consumption habits in France lead to
competition by lowering prices more than competition by increasing quality and providing
additional services.
Looking at possible future intervention in labour costs, Cadrey and Jany-Catrice estimate that
four channels must be considered before arriving at a ‘pure’ increase in work expenditures.
The first is the increase in trade activity (volume of products sold), the second is work sharing
by the extension of part-time (constrained, involuntary part-time, they fear), the third is the
substitution of employed workers for self-employed workers (because big firms can eliminate
116
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
some small shops by using the increased margins to lower prices), the fourth is a sectoral shift;
some retail trade firms could internalise warehouse functions to gain better control of the
supply and reap economies of scale. Lastly, they consider the increase of work used; they see
this effect as negligible in the French case.
Finally, Gadrey and Jany-Catrice enlarge the debate by introducing the case of Japan,
Germany and the United Kingdom. They insist on the interaction of four ‘socioeconomic
spaces,’ namely the market competition structure, the labour law and industrial relations
space, the firm’s organisational choices, and the domestic division of labour. Within such a
framework, the wage costs should be considered to interact with service quality but also with
the distribution of family and outside working time as well as the productivity choices. Then
the ‘social habits’ appear as a global path-dependent process, opening the way to persisting
‘societal’ differences.
4.2.3. Assessing the controversy
Some methodological remarks are in order here. First, neither Piketty nor Gadrey and JanyCatrice present a complete set of arguments about the underlying causality, checked by
econometric testing. It must be acknowledged that such a task would be delicate, notably
because one needs to test numerous relationships.
Second, their time perspective is different. Piketty insists on very long-term (secular)
tendencies, and this allows him to suggest that the consumption structures and levels should
converge if the price and cost levels are flexible enough. Interestingly, he states that the French
consumers should be glad to benefit from the additional selling services offered to the United
States consumers for the same price, and is sceptical about supposed ‘social habits’. The
question of the competition structure in the sector is not considered, because in the very longterm the consumer would win and impose quality improvements. It is a view that directly
combines cost distortions and consumer pressures, and ignores more or less intermediary
causalities.
Gadrey and Jany-Catrice present a mid-term perspective. They argue that it is not realistic to
speculate on a quick convergence of consumption levels, and that the same applies to the
consumption structure: some additional services could be welcomed by the French
consumers, but not all (2). Cutting the social security taxes, in their view, would simply lead
to a lowering of the share of personnel costs in the total costs, not to new employment.
The debate is then concentrated on a possible threshold effect; for Gadrey and Jany-Catrice
the absence of real response by the French retail trade sector to social security cuts (centred
on low wages) implemented during the 1990s, shows that there is a perverse windfall effect
unless there are other painful and complicated changes, at least in competition and family
(2) They admit that more information and help could be welcome, but feel that ‘open 24 hours’ does not correspond
to any demand in the French case.
117
The job creation potential of the service sector in Europe
structures. For Piketty there is mainly a question of intensity and persistence of wage cost cuts;
he observes that the adaptation of firms requires stabilised long-term expectations, and states
that the social security cuts in France during the 1990s remained limited and were not
perceived as durable.
He presents the case of French domestic home services as a confirmation of this. After a few
years of heavy subsidisation (during the first half of the 1990s), favourable to high-level
households allowed to deduce part of their expenditures from their tax bill together with social
security tax cuts, the results were impressive and led to a considerable increase in domestic
employment (of course, partly due to the official appearance of hidden employment). The
maximum subsidy could amount to 50 to 80 % of the total wage cost. According to two
different evaluations, the number of employees doubled between 1990 and 1996 or witnessed
more than a 50 % increase (Piketty p. 87) (3). Here, the threshold was seemingly overcome by
the mere size of a subsidy much bigger than in the general case of low wages.
There was no resistance, however, and Piketty admits a French preference for domestic
employment when compared with the United States. Gadrey and Jany-Catrice argue that in
fact the real cost cut in the case of retail trade was bigger than that stated by Piketty, because
it combined a general cut on the low wages’ social security contribution with another cut on
part-time work. So they contend, for the 1993–98 period, that the social security contributions
paid for a part-time cashier were at the same level in France and the United States (4).
Nonetheless, a complementary fact not fully discussed by the three authors can be taken into
account here. It is the quick response of the French firms (and French workers, either
voluntarily or involuntarily) to the heavy subsidisation of part-time work during the second
half of the 1990s. The share of part-time work in retail trade jumped from 16.4 % in 1992 to
21.3 % in 1997. It reflects a more general (nation-wide) trend.
The price response here seems very clear, and it appears to have easily overcome, maybe with
important drawbacks, what Gadrey and Jany-Catrice call, surprisingly, the French ‘resistance
to part-time work’ (ibid., p. 115). They observed that the retail trade sector lost jobs during the
1992–97 period (ibid., Table 1 p. 106).
However, the essential point is that labour costs cuts lead to highly differentiated responses,
that cannot be reduced to ‘social resistances’, either arbitrary and exogenous (Piketty), or
homogeneous, closely intricated and stabilised (Gadrey and Jany-Catrice). Other thresholds
must be introduced in the debate, regarding for example, the modalities of work and work
sharing. It is especially the case in the retail trade, where low-skill female employment is of
paramount importance. Work composition and work volume responses seem to be partly
independent, but also connected in an overall cost and organisational structure. We find here,
(3) The principle of giving such advantages to rich taxpayers was publicly criticized and at the end of the 1990s
the French Government decided to decrease the subsidies.
(4) Here again the advantages given were felt excessive by the government responsible, and since 1998 there has
been no particular advantage in developing part-time work.
118
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
at a sector’s level, the whole classical span of firms’ and workers’ responses to a wage subsidy,
including windfall and substitution effects.
How then shall we identify the range of possible increases and substitutions, and their
determinants? Two elements seem useful in order to develop international comparisons
without falling into general ‘societal’ considerations.
— First, introducing the idea of ‘functional equivalents’, that is, different sets of production
and arrangements corresponding to a given set of services or consumption. These
‘functional equivalents’ do not need to be perfect equivalents. Indeed, it is quite the
contrary; from one country to another, from one social group or one income level group to
another, the needs can be (more or less) satisfied in different manners. For example, child
rearing could be organised on a family basis, on an external service basis, or on an
intermediary basis. The point here is to take into account the possible spillover effects
stemming from a given arrangement and to introduce them in the comparison.
— Second, there is a need to limit oneself to an intermediary level between pure cost levels
and societal interlocked global causal networks.
The following section will discuss previous developments in the risk management and
organisational choices underlying work relations (5).
The labour law regime of jobs (in a broad sense, including social security arrangements and
public employment policy), seems to be of prominent importance for the employers as well as
for the employees. It involves the distribution of rights and risks together with costs and
incomes. It combines the effects of the social protection regime and the professional choices
of the firms and the individuals. It can be modified, when taking into account the connection
with other arrangements. Path dependency effects, which obviously exist, can be
differentiated.
Some of the choices seem to rely on the interplay of precise institutional sets and
organisational choices.
We saw that the retail trade sector in France is dominated by a small number of highly
concentrated capital-intensive firms, which mainly use a low-skilled female workforce and
have the permanent goal of lowering all costs. In such a context, the switch to labour-intensive
behaviour would be equivalent to a change in the competition behaviour model because it
implies the production of additional services, namely, introducing more client-oriented
attitudes and more quality in the distribution process.
This may indirectly and paradoxically correspond to important changes in actual employment
patterns: we saw in the ‘service profit chain’ (6) that the client-oriented attitude is fostered by
(5) Cf. the contribution on business services employment and externalisation by Gazier and Thévenot, in this
volume.
(6) Cf. Gazier and Thévenot, op. cit. One important example given in the book by Heskett, Sasser and Schlesinger
is the case of Wal-Mart, the big United States department store.
119
The job creation potential of the service sector in Europe
a stabilised and rather well-paid, autonomous workforce. The risk sharing choices may
strongly differ between the employment and compensation of a low-paid minimal sales force,
and that of a client-minded sales force, which must be more autonomous and somehow
associated with the gains. Developing additional services leads, at least, to diversifying the
workforce and securing some skilled careers, instead of massively using the actual labour
cost-lowering policy schemes (social security tax cuts, part-time incentives). Such an
evolution, while possible and desirable, would take an extremely long time. Thus, the
employment regime may be an intermediary level between labour costs and social habits, an
interesting level of inquiry, explanation, and intervention.
4.3. Employment in the performing arts and audio-visual activities in
France and the United Kingdom
The growth of cultural employment in France and in the United Kingdom during the period
1981–92 is impressive. Nevertheless, it reflects a twofold trend; an increase in the volume of
employment as well as structural changes, thanks to the policy carried out in the two countries.
After a brief summary of the context of the research, we analyse the convergences and
divergences between the trends of employment in France and the United Kingdom. The
interpretation of the issues leads to hypotheses concerning the determinants of the respective
evolution of the volume of employment in France and in the United Kingdom.
4.3.1. Employment in cultural activities and Baumol’s cost disease
Characterising all economic activity as either stagnant or progressive, Baumol and Bowen
developed a model of unbalanced growth in 1965 which predicts the increasing relative costs
of providing certain service industries. The idea has its origin in the empirical study of
performing arts which faced a deep crisis on Broadway in the 1960s (7).
This model gave birth to a huge body of economic literature which tested its empirical
conclusions by measuring the differential growth rate of prices in the performing arts and in
the general economy. Few studies discuss the hypothesis concerning employment:
— the productivity improvements in the arts lag behind those of the entire economy;
— wage increases in the arts have to keep up with those in the general economy.
Cultural employment then shares the general characteristics of service employment,
especially the weak distinction between the product of the work and the work itself. However,
some specificities of cultural employment may justify a focus on the way regulations
intervene in order to organise job relations.
(7) For details concerning Baumol’ s work in the cultural field, cf. Towse (ed.), 1997.
120
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
We will not discuss the limits of the productivity lag argument, especially if we consider the
performing arts as a stage in the production process of audio-visual services. A large part of
the performing arts activities is used as a central means for the promotion of cultural industries
(records, videos, etc.). According to Schmid (this volume), Baumol’s cost disease can be
overcome through service product chains including broadcasting, recording and
merchandising.
The question of wages has been studied in a different framework than Baumol’s. Many
contributions discuss the rationale of entering artistic careers (Menger, 1989) by testing the
average earnings of artists compared with earnings of non-artists with an equivalent level of
qualification, and also by modelling the choice between artistic and non-artistic careers
(Singer, 1981). They conclude that there is a low wage differential between them and point to
the ‘myth of the starving artist’ (Filer, 1986). From this point of view the results do not refute
Baumol’s conclusions.
While researchers discuss the determinants of the choice between artistic and non-artistic
careers, many studies describe cultural employment markets as atypical and characterised by
a strong and unavoidable flexibility (Towse, 1992 and 1993).
In this section we focus on the respective forms of this flexibility in France and in the United
Kingdom and on the factors that have favoured their development.
4.3.2. The growth in cultural employment: Some convergence between France and the
United Kingdom (8)
4.3.2.1. Convergences
➢ Illusional job markets?
The growth of cultural employment is apparently strong, compared with the increase in the
general workforce.
— The growth of cultural employment is 16 % between 1981 and 1991 in the United
Kingdom, while the growth of the active workforce is negligible, and 42 % in France
between 1981 and 1992, while the growth of general employment is 3.7 % (O’Brien and
Feist, 1995: 97, Observatoire de l’emploi culturel, 1993). During the same period the
growth of employment is stronger for the United Kingdom if only artistic or cultural
professions are evaluated; 34 %, including cultural occupations outside the cultural sector.
In France, individuals with artistic and cultural professions increased as well by 37 %
between 1982 and 1992, without taking into account the growth of cultural occupations
outside the cultural sector (id.).
(8) A different version of this paper is forthcoming in the journal of cultural economics.
121
The job creation potential of the service sector in Europe
— Among the cultural industries, the most impressive growth rates concern audio-visual and
performing arts in both countries. In France, they reached 90.2 % for the performing arts,
and 67.5 % for the radio and television activities. The only sector which faced a decrease
in employment (about 30 %) is film distribution and movie theatre management, resulting
from a growing concentration of these activities, while the increase in the production was
high, 122.4 % (Observatoire de l’emploi culturel, 1993). In the United Kingdom, the
growth rates of employment in film, radio and television activities vary from 19 to 25 %,
and are much higher when taking into account the self-employed artists (O’Brien and
Feist, 1995).
— Non-artistic jobs (such as ticket sellers) are complementary jobs linked to the growth in
artistic production (Cabinet Ithaque, Cabinet Themis 1997). If we take into account only
the individuals with cultural occupations, instead of the total number of employees
working in the performing arts and audio-visual activities in France, the growth rates vary;
between 1982 and 1990, from 37 % (musicians and singers) to 123 % (pop music artists),
with an average rate of 47.6 % (Observatoire de l’emploi culturel, 1993). In the United
Kingdom, among cultural professions, stronger growth (47 %) is among artists, actors,
entertainers, singers, and stage managers (O’ Brien and Feist, 1997).
— However, the volume of employment increased more slowly than the number of
individuals in both countries. For example, in France the number of technicians in the
performing arts increased by 91 % between 1986 and 1991, while the number of wage
earning contracts grew almost 300 % (Ministère de la Culture, 1995). Between 1986 and
1994, the number of actors increased from 6 000 to 12 000 while the work hours decreased
25 % (Menger, 1997). Thus job opportunities seem high, resulting from the increase in the
volume of employment as well as from the illusion which results from the shortening of
contracts, through a process of work sharing.
Moreover, many new entrants overestimate their chances, then gain information and
eventually leave the career. Only half of those with cultural occupations in 1981 were in
cultural occupations ten years later in the United Kingdom (O’Brien and Feist 1997).
Cultural markets can be viewed as illusional markets, thanks to the shortening length of
contracts and to the number of entries compared with the duration of careers. The high level
of rewards for those who succeed is an incentive for entering the market. Large dispersion of
rewards, extreme concentration at the top of the heap, and a low probability of obtaining high
remuneration make cultural markets a kind of lottery (Rosen, 1986).
➢ A high degree of flexibility
As project by project employment tends to predominate, reflecting an increasing rhythm of
obsolescence and the need for constant innovation (Stinchcombe, 1968, Benhamou, 1997) a
high degree of flexibility developed much earlier than in other service industries.
— The development of adhocracies, defined as project by project organisations (Mintzberg,
1979), is obvious in the performing arts. In France about 50 % of the 585 festivals existing
in 1995 were created between 1980 and 1995 (Benito 1995).
122
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
— Even in the public cultural administration one can observe a decrease in the number of
permanent employees. The Centre Georges Pompidou in Paris employs 1 500 workers and
among them only 832 have permanent job contracts; 120 are hired from time to time,
without any regular part-time contract. At the Bibliothèque de France, 540 non-permanent
workers are employed versus 2 500 permanent salaried workers (Le Monde, 4 February
1999).
— The number of freelance workers has grown in film production (Christopherson and
Storper, 1989, Nicolas, 1995), while the length of wage earning workers’ contracts was
drastically reduced.
— In the television industries, permanent wage earning employment has fallen while flexible
employment increased, thanks to changes in regulations and in the structure of the
production (Dex, 1998). Between 1988, 1992, and 1996, permanent employment in the
public television sector decreased from 23 206 to 21 981 and 16 681 units in the United
Kingdom. It decreased from 4 750 to 4 650 and again 4 750 in France (Conseil de
l’Europe, 1996), while the production rose with the multiplication of networks.
— In the United Kingdom, incentives and regulations encouraged the rise in selfemployment such as:
• The BBC ‘producer’s choice’ initiative encouraged the trend in the media sector by
contracting out in small independent companies (Woolf and Holly, 1994, O’Brien and
Feist 1997).
• Following a proposal of the committee on the finance of broadcasting (9) (1985–86) the
government in the late 1980s decreed that 25 % of TV programming should be bought
from ‘independent’ producers.
The BBC encouraged its own producers to become independent. More generally, different
structural changes (competition with private sector broadcasters, the creation of Channel 4,
the reorganisation of the company from vertically divided directorates into horizontally
divided sections (10)) led to an increase in the proportion of freelance workers (Benhamou,
forthcoming).
In such an organisational context, internal job markets (Doeringer and Piore, 1971) (11) are
rare. The discontinuity of relations between workers and employers concerns the whole job
market. Even in the primary job market (12) high qualifications (13) go hand in hand with a high
(9) The committee made the proposal that 40 % of TV programming should be bought from ‘independent’
producers.
(10) For more details, see Dex et al.,1998.
(11) We could make an analogy with the level of wages in sport competition as Bouvet suggests (1996).
(12) Economists often distinguish two partially unified labour markets (Doeringer and Piore 1971). Some
professions do not fit this model, like blue-collar craftsmen, who are highly mobile and highly skilled (Okun,
1981).
(13) In France, 55 % of the individuals with cultural occupations have a higher education level than the ‘bac’;
32 % have a high degree diploma; 25 % no ‘bac’ (Observatoire de l’emploi culturel 1993). In 1991, in the
United Kingdom, 76 % of the individuals in the total work force had educational or training qualifications
compared with 93 % of those in the cultural activities (O’Brien and Feist, 1997).
123
The job creation potential of the service sector in Europe
turnover rate and a weak job security. Though it is a factor of rising wages, hiring well-known
artists or highly skilled technicians over a short period cuts down the risk associated with
competition and sends quality signals to the consumer (Rosen 1981).
The quasi-absence of internal job markets means that the constitution of a solid reputation is
central for job candidates, who develop multi-activity in order to face periods without any
cultural occupation. Many workers have several part-time jobs in and outside the arts sector
(Towse, 1992, Greffe, 1996). In France, in 1995, 10 % of the individuals who had a cultural
profession declared that they had another activity during the same period, but 3 % for the
whole economically active population, and 18 % among artists (Observatoire de l’emploi
culturel, 1996). In the United Kingdom, Peacock and Weir (1975) notice that musicians
combine different job contracts to face the discontinuity of their activity. In 1991, 5 % of the
general workforce had a second job compared with an average of 9 % of those working at
cultural occupations and 14 % of those working as musicians (O’Brien and Feist, 1997). For
this they are aided by relatively high qualifications and reconversion capacity. Thus
employability, which consists of an accumulation of human capital (skills, reputation,
networks) which can be invested in new job opportunities (Moss Kanter, 1995), becomes a
substitute for job security.
As mobility reigns, the worker builds a reputation by moving from one project to another.
Workers who rotate through different workplaces and jobs have a greater network of
connections with people who have jobs to offer (Ferber and Waldfogel, 1996). This
phenomenon is a characteristic of software industry workers, who need to make sure that their
knowledge is transferable. They link their careers to the whole industry and do not identify
their job with a unique project or firm: ‘Changing employers is so common that professionals
may identify their interests more with the project and the industry than with the particular firm
where they happen to be.’ (Moss Kanter, 1995).
4.3.2.2. Divergences
Though detailed employment data are to be interpreted with much caution (14) (see appendix),
a comparison between France and the United Kingdom leads to the following issues.
➢ An inverse trend in the share of part-time jobs
In the United Kingdom, the proportion of part-time jobs (15) increased in the economy as a
whole, while it decreased for wage-earning female workers though the number of job
creations went up. For men, the percentage of full-time workers remained stable or decreased
slightly in the radio, television and theatre sectors (from 91 to 83 % between 1981 and 1991,
see O’Brien and Feist, 1995), but at a very high level compared with the situation in France.
(14) See O’Reilly, 1996.
(15) Part-time includes shorter working time than the full-time volume of hours, and also temporary jobs which may
be analysed as part-time jobs on a yearly basis (Rogowski and Schöman, 1996).
124
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
There the number of actors grew by 10 to 20 % from 1986 to 1991, while the annual number
of working days decreased at an average rate of 5 % (and 14 % between 1989 and 1990, see
Menger, 1997).
In France, intermittent employment (wage-earners with fixed-term contracts) is more and
more frequent in the performing arts and audio-visual sectors. Among technicians, 30 000 jobs
were granted by 6 700 companies in 1991. The length of job contracts shortened starting in
1985. Their average length diminished from 32.5 to 13.5 days between 1985 and 1991, in spite
of the growth in demand for employment from the companies. The growth rate of the number
of working days was 62 % during the same period while the number of contracts increased at
a rate of 300 % (Ministère de la Culture, 1995). Between 1986 and 1991 the average number
of working hours for actors decreased (16) while their total number grew, leading to a rise in
part-time employment.
➢ Self-employment: an increase much greater in the United Kingdom than in France
In the United Kingdom, 34 % of individuals with cultural occupations were self- employed (17)
in 1994, compared with 13 % for the total economically active population (18) (Cultural
Trends, 1995). Between 1981 and 1991 there was a 72 % increase in the number of selfemployed individuals in the cultural industries — while the number of employees (full and
part-time) increased by 7 %. (O’Brien and Feist, 1995) (19). Moreover, there is evidence of a
significant shift from employed to self-employed status. Nearly two out of five musicians who
were employees in 1981 were self-employed in 1991, as were 23 % of actors, entertainers, etc.
Overall, some 15 % of those who were employees in 1981 were self-employed in 1991
(O’Brien and Feist, 1997). These tendencies have recently been reinforced; 28 % of
individuals in films and 67 % in the performing arts were self-employed in 1994, while the
figure was 13 % for the total economically active population (Casey et al., 1995). Entrants in
those sectors in 1991 were, respectively, 40 % full-time employed, 6 % part-time, 33 % selfemployed, 10 % waiting to take jobs or unemployed, and 11 % with diverse statuses (O’Brien
and Feist, 1997).
(16) Number of actors and annual number of working days, annual variation, France, 1986–91 (%).
Annual growth rate of:
Number of actors
Annual number of working days
1987/86
15
–4
1988/87
13
–6
1989/88
9
1
1990/89
18
– 14
1991/90
10
–5
Source: Menger 1997.
(17) The self-employed are those who work for their own benefit rather than for an employer in a conventional
(dependent) employment relationship (Meger 1994: 184).
(18) However, 28 % in the film industry, 67 % in the visual, literature and performing arts in 1994 (Cultural Trends,
1995, 25).
(19) In the United Kingdom, between 1981 and 1991, the number of self-employed grew 96 % in film production,
distribution and exhibition (vs 13 % for all employees), 123 % in radio, TV services, and theatres (vs 14 % for
all employees), and 64 % among authors, music composers and other artists not elsewhere specified (vs 60 %
for all employees) (O’Brien and Feist, 1995).
125
The job creation potential of the service sector in Europe
In France, on the contrary, for individuals in 1995 with a cultural occupation, the selfemployment rate was lower than for the total economic workforce (17 %) in the film
industries, in the radio and television services (11.5 and 7.7 %), and higher in the performing
arts (32 %) (20), but much lower than in the United Kingdom. Moreover, these figures are
diminished when taking into account only the cultural professionals; 15.5 % for the audiovisual and performing arts activities in 1995 (Observatoire de l’Emploi culturel, 1996).
4.3.3. Toward an interpretation of the divergences between France and the United
Kingdom: A specific way of managing risk
4.3.3.1. Two models of market arrangement
Cultural economics focuses on the search for innovation and on the growing propensity of
firms to produce ephemeral goods and services. For this, firms need to hire specific
qualifications and the constitution of ad hoc teams. The rhythm of obsolescence of the
production implies the management of specific risks. Working on temporary projects implies
the creation of non-standard jobs at the expense of regular workers, even when production is
rising.
Williamson (1975) shows that the need for specific assets should lead to hierarchical
solutions. However, in the case of cultural production this need is temporary and linked to
short-term projects. Thus, it is a case of internal spot markets, and we can make the hypothesis
that this permanent search for specific short-term commitments implies a market solution
instead of a hierarchical one. Two different models of labour market adjustments may be
adopted in order to apply such a solution.
— In France, temporary employment is a way of avoiding a permanent hierarchical solution,
without a total market arrangement. This hybrid solution has a cost, which is borne by
workers inside and outside the cultural sector (see note, infra).
— In contrast, according to an apparent improvement of the cultural job market, part-time
employment, which is most vulnerable, decreased in the United Kingdom. Selfemployment led firms to lower labour costs, especially the costs resulting from the
discontinuity of production activity (irregular earnings, costs of contract seeking).
In both cases, the strategy of firms consisted of externalising costs through the creation of
unemployment. The final effect was to create a pool of applicants by giving the illusion of a
strong increased rhythm of job creation and of low entry barriers. The illusion is stronger with
the ‘overwhelming conceit’ with which entrants were said to assess their chances of
success (21).
(20) France exhibited a slight increase in self-employment in the economy as a whole during the 1980s (Meager,
1994).
(21) Smith (1947) quoted in Rosen (1986: 681).
126
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
4.3.3.2. The contribution of policies for the development of part-time and/or self-employment
jobs
Legal labour regulations have shaped job relations in both countries. The regulatory context
has contributed to the growth in self-employment in one case and to the growth in fixed-term
job contracts in the other (22). The degree of flexibility introduced by legal reforms and the
extent of recourse to non-standard employment appears to be closely linked, even if not easy
to evaluate.
➢ In the United Kingdom
Firstly, the general legal and institutional environment is favourable to the flexibility of work;
job market regulations are weaker than in France. Cultural employment has benefited from the
flexibility of the employment market of the whole economy since the 1970s (O’Reilly, 1996).
The practice of ‘lean production’ was adopted earlier than in France, and the growth of selfemployment was particularly marked for the whole economy from the early 1980s on (Meager
and Bates, 1997). Between 1979 and 1989 it almost doubled (Campbell and Daly, 1992).
Business start-up is easier and capital market rather less strict than in other members of the
European Union (Meager, 1994). In the same way, the small increase in temporary job
contracts could have resulted from weak regulations of the labour market and easy dismissal
procedures (Mosley, 1994).
Secondly, though the EAS programme (Enterprise Allowance Scheme) (23) introduced in 1983
does not concern the cultural sector specifically, it helped the unemployed to switch to selfemployment. Assistance was in the form of a weekly allowance offered in the first year of setup. Between 1985 and 1986 and 1989 and 1990, the total number of individuals participating
in EAS with a cultural sector business or occupation was 35 177 (Feist and Eckstein, 1991, in
O’Brien and Feist, 1995).
Part of the newly self-employed were previously employees working inside the companies,
whose jobs, through new contractual arrangements, were externalised by companies in order
to lower costs (24) (see also Gazier and Thevenot, this volume). Apart from this process of
(22) The contrasted development of part-time in the two countries is surely related to the different incidence of parttime; nevertheless there was no catching-up in the United Kingdom, and it remained relatively low during the
period.
(23) Which does not only concern the cultural sphere. However, there has been an over-representation of service
sector activities among EAS participants because of the weakness in the eligibility conditions and the low entry
barriers (Meager, 1994).
(24) The EAS programme was also a factor in the improvement of employability: ‘… even if the enterprises started
under the scheme are unsuccessful, and even if the participants leave self-employment after (or during) the
subsidy period, the skills and experience gained through business start-up will stand them in good stead in the
labour market, improving their attractiveness to potential employers’ (Meager, 1996: 496). An evaluation of
the employability effect would require longitudinal data that are not available. United States pilot studies have
tried to evaluate it (Benus, Wood and Grover 1994) and conclude that there is an improvement of employability
through support for self-employment.
127
The job creation potential of the service sector in Europe
exogenous flexibility, a kind of ‘natural’, endogenous flexibility is a permanent characteristic
of cultural markets, which found a source of public funds in the EAS programme.
➢ In France
Cultural labour markets developed within a specific regulatory context, in contrast with the
relatively weak degree of deregulation of the labour market in general between 1980 and
1991. Financial incentives have encouraged part-time and temporary work through the social
security system.
Since 1969, a specific scheme for the performing arts and the film industry has compensated
unemployment periods in the performing and audio-visual activities for artists and
technicians, under certain conditions (see box 1). This system has been deeply criticised:
— Firstly, because it works as an insurance system and does not compensate for inequalities.
— Secondly, because of its cost.
— Thirdly, because the cost is not borne by firms and workers in the sector but by the whole
unemployment insurance fund for salaried workers which absorbs the increasing deficits
of the particular performing arts unemployment insurance scheme.
— Fourthly, the system is supposed to generate adaptive behaviours. On one hand, firms treat
unemployment benefits as an indirect means of subsidising production by lowering labour
costs. They create unemployment when a project is over. The worker bears the risk but
transfers it to the whole social security system. On the other hand, workers build networks
by moving from one job to another.
Intermittency (25) increases as a result. For example, in 1990 there were 150 000 wage earning
workers in the audio-visual and performing arts field. Among them were 96 000 intermittent
workers in 17 000 firms, with job contracts of shorter duration. In addition 20 000 firms had
declared less than six job contracts during the year (Observatoire de l’emploi culturel, 1994).
In 1990, in the audio-visual and the performing arts sectors:
— 52 000 artists signed 423 500 job contracts, which represented 3 383 000 working days,
with an average length of one job contract each of eight days.
— 44 000 technicians signed 178 000 job contracts, which represented 4 000 000 working
days, with an average length of one job contract each of 22 days (GRISS and Observatoire
de l’emploi culturel, 1994).
Intermittent workers earn more during short time periods than regular employees on average,
therefore a private compensation is added to the social compensation for the risk linked to
(25) In the case of intermittency, teams are assembled for particular ventures, and job security is based on reputation
and networks.
128
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
fixed-term contracts. Job insecurity costs are internalised by the market through higher
earnings (26). ‘… if prospects for success in some fields are uncertain, the mean earnings among
practitioners must adjust to attract risk-adverse applicants. Hence, occupations in which there
is greater uncertainty about possible income must exhibit larger mean earnings than those in
which a modicum is virtually guaranteed’ (Rosen, 1986: 681).
Box 1: The performing arts unemployment insurance scheme
Workers who have been working for 507 hours in the last 12 months can benefit from the
unemployment insurance scheme to compensate for the absence of any earnings during the
periods of inactivity. The amount of money paid during this period is a percentage of the
earnings of the artist. For more details, see Paradeise, 1998.
About 80 000 intermittent workers have paid a contribution of EUR 91.5 million in 1997
and they have receive EUR 550 million during the same period. The deficit is absorbed by
full-time salaried workers.
France is re-examining this protection system, because of its cost and of the decreasing
specificity of cultural markets. Hyperflexibility becomes a rising model for the whole
employment market.
4.4. Job creation and flexibility in cultural activities: Some hypotheses
One cannot conclude in favour of one model over the other for job creation. However, a few
hypotheses and issues emerge from this research:
1. The increase in the volume of employment is much higher in France, especially within
cultural activities. The growth rates were 42 % in France, and 16 % in the United Kingdom,
much higher than in the general economy. This difference disappears when we take into
account the growth of cultural professions outside the cultural activities. How can we
interpret these figures?
Firstly, in both cases, facilitating flexibility has been a condition for the growth of
employment. Secondly, probable disparities in the output volume have contributed to the
differential in the growth rates of employment in the two countries; though cultural activities
are heterogeneous, further research could lead to an evaluation of the respective growth of this
volume in France and in the United Kingdom during the period. Thirdly, cultural employment
takes place very often in non-profit institutions, whose funds depend on private and public
spending. The fluctuations of public cultural funding bring about fluctuations in the
production of cultural goods and services by these institutions (Greffe, 1996). However, the
changes of the public commitment to the arts and culture does not necessarily imply identical
(26) For more details, see Gurgand and Menger, 1996.
129
The job creation potential of the service sector in Europe
fluctuations of the growth of cultural employment. Employment cannot be mechanically
linked to the production level of cultural goods and services.
Nevertheless, public spending for the arts has been relatively weak in the United Kingdom
during the period, compared with the commitment of central and local authorities in France.
We can hypothesise that this commitment has favoured employment growth in France. This
hypothesis is reinforced by the large proportion of non-profit public institutions in this
country. Further research should test this hypothesis, and combine case studies and
macroeconomic data analysis.
2. The French policy based on social security benefits is most likely exhausting its
resources (27). A decrease in these indirect subsidies would probably lead to a decrease in
the growth rate of cultural employment. If government wishes to promote cultural
employment, an analysis of a stronger support policy towards the demand for culture, and
of its short- and long-term results should be considered.
3. Some researchers have tested the hypothesis that the specific insurance system would have
given workers the incentive of multiplying short-term assignments in France (Gurgand and
Menger, 1996). By only comparing the United Kingdom and France, this hypothesis
cannot be evaluated. Nevertheless, the high growth in part-time employment could be a
symptom of the effects of social security on the behaviour of both workers and firms,
through a learning process which would have led them to the use of social benefits as
indirect subsidies for their own cultural occupations and activities.
4. Moreover, the comparison of the individual and social costs of the ‘hyperflexibility’
inherent in cultural activity should be made in both countries. Such research could help
governments adjust regulations and social security systems to their employment objectives
(namely, growth in the volume of employment versus decrease in part-time employment).
In order to answer the need for ‘hyperflexibility’ in cultural activities, especially in the
performing arts and audio-visual activities, different models of market adjustment can be
adopted. The flexibility regimes analysed in this chapter are linked to the institutional and
societal specificities of the welfare state regimes. On one hand, in the United Kingdom, the
liberal regime means that cultural activities benefit from general employment policy and
assume a deregulated context for job relations, leading to growth in self-employment. On the
other hand, the increase in employment in France is partially due to a corporatist continental
welfare state regime.
(27) See Paradeise, 1998.
130
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
4.5. Conclusions and policy recommendations
The focus of this chapter is the importance of considering ‘employment regimes’ and the
structuring of markets in sectoral international comparisons of employment performances.
The point is made twice; once for retail trade and once for cultural employment.
For retail trade, we showed in our assessment of the France/United States gap that
explanations based on simple labour costs or ‘societal habits’ are not fully convincing, and that
the competition structure and employment patterns of the sector, combined with the incidence
of public employment policies, are useful elements to introduce in the discussion. Cultural
activities on the other hand are characterised in our comparison field (France/United
Kingdom) by a strong increase in employment, accompanied by a high degree of work
flexibility. Nevertheless, the comparison shows that different models of flexibility can be
adopted according to the national specificities of the welfare state regimes. These specificities
influence the growth in the volume of employment and lead to a process of work sharing. A
survey of the case of a rather liberal system (the United Kingdom) and a more interventionist
regime (France) leads to a twofold issue. Firstly, the increase in the volume of employment is
relatively indifferent to the regime, secondly, in both cases, the costs of the flexibility are not
borne by the same economic agents, and they may threaten the future of the national regimes.
Recommendations
Four main policy consequences stem from the comparisons made which stress the importance
and idea of ‘employment regimes’.
The first is that lowering labour costs as a policy cannot be considered in isolation, and should
be related to the interplay of labour market institutions and family constraints and or
opportunities. The policy choices concern a wider set of arrangements as well as subsidies or
other cost-lowering devices.
The second is that activity volume, though a key parameter, is virtually unknown. In the case
of retail trade, the difference between France and the United States is mainly explained by a
difference in volumes sold, and by the presence and or absence of some additional selling
services. In the cases of retail trade and cultural activities, it is useful to estimate more closely
the trend in the volume of outputs in a long-term perspective in order to better understand and
estimate the link between the increase in this volume and the increase in employment.
The third consequence, which is actually common sense, is the need to consider both sides of
the market as well as the competition structure when trying to develop employment in a given
sector; a unilateral ‘supply’ or ‘demand’ approach may lead to inefficient results. The case of
retail trade is a good illustration of the importance of firms’ sizes and strategies. The
perspective of increasing the quality of the workforce, through training and quality-promoting
institutions, makes some sense. In the case of cultural employment, it seems advisable to
extend the different means for increasing the flexibility of employment and the diversity of
firms (which are a condition for the growth in employment in the cultural activities) through
131
The job creation potential of the service sector in Europe
loans for set-up, tax and social contribution deductions and time extensions for paying taxes.
One may also strengthen the roll of non-profit institutions in these activities, by promoting
incentives for these organisations to encourage both job creation and product diversity.
Finally, risk-management in and around work relations may lead to different solutions. These
may seem to be equivalent in the short run but may reveal functional differences in the long
run, when compensating arrangements, dependent on other institutions, exhibit limits or
adaptability. The real social cost and feasibility of a given solution may only appear on an
enlarged basis, as is clear in the debate about cultural employment, which includes the choice
of subsidising cultural activities.
Appendix
Data on cultural employment: Sources and interpretation
1. Sources include public administration, professional organisations, the social security
administration, and the censuses. The most reliable source is the census of the population
because of its size, coverage and response rate (completion of the forms is compulsory).
Censuses of the population were conducted in France in 1982 and 1990 (and 1999), and in
the United Kingdom in 1981 and 1991. They provide quantitative and qualitative elements
(demographic characteristics, economic status, regional distribution) that allow a
comparison of the evolution of employment in both countries. They may be completed
with issues of longitudinal studies covering cultural employment (28) over the same period
which provide useful qualitative information.
2. Changes in the methods and the nomenclature may occur, which prevent comparisons
through time. For example, in France, the growth between 1990 and 1995 should be
studied with caution because of the change of the nomenclature which occurred in 1993.
Among different changes, most art-teaching activities and different leisure activities are
excluded.
3. Moreover:
— Some cultural activities cannot be isolated because of their narrow scope.
— Cultural workers often accumulate several activities inside or outside the cultural
sphere. In both countries, the main activity is taken into account, though the
interpretation of the main activity may vary from one individual to another. A large part
of the uncertainty remains because of the subjectivity of workers concerning the
definition of their main activity (29).
(28) This study covers five standard occupational classifications: authors, writers, journalists (SOC 380), artists,
commercial artists, graphic designers (SOC 381), actors, entertainers, stage managers, producers and directors
(SOC 384), musicians (SOC 385), photographers, camera, sound, video equipment operators (SOC 386).
(29) The growth in the number of art teachers reflects different complementary phenomena; teaching is either a
second job, a main job, or a job for unsuccessful artists. Data do not provide strong qualitative elements in order
to evaluate the proportion of each case.
132
Labour costs, social habits and employment regimes. Two comparisons of national employment
performances and some lessons
4. The two countries have adopted two different options in order to evaluate the growth of
employment. In this study we analyse both methods.
The activity of the economic unit where the job takes place — artistic and non-artistic
employment are both evaluated each time they take place inside the cultural sectors.
The profession (the nature of the job), which takes into account only artistic and cultural jobs.
The institutional framework of cultural activities is rather different in the two countries. The
French tradition of isolating cultural activities can be opposed to the British propensity for
considering cultural activities in their economic dimension as other economic activities, with
their own specificities. Thus, France provides detailed statistics on the cultural labour market
structure, while in the United Kingdom, cultural statistics include cultural professionals
outside the cultural sphere. Statistics reflect both institutional and organisational structures.
133
Part II:
The industry and service nexus:
Concepts and empirical evidence
Chapter 1
The nature and determinants of service
sector growth in the United States
Randall Eberts and George Erickcek
1.1. Introduction
Most developed economies have experienced significant structural change during the last
several decades, as service jobs have supplanted manufacturing jobs. In the United States, the
creation of service jobs has fuelled most of its historically high employment growth, while
growth in manufacturing jobs has remained flat. The European Union (EU) has also
experienced significant growth in service jobs, but its pace remains slightly behind that of the
United States. What concerns European policy-makers is that service sector jobs have not
grown fast enough to close the gap in the employment share of services between the EU and
the United States. Although the service sector’s share of total EU employment has increased
from 48 % in 1974 to 65 % in 1997, its share still falls short of the United States share of
73 %. Of equal concern is the relatively slow growth in total employment in the EU. Unlike
the experience in the United States over the last two decades, total employment in the EU has
grown at only half the rate of that in the United States. Based on the United States experience,
policy-makers have increasingly looked to the service sector as the catalyst for greater growth
in total employment. Yet, while the employment growth potential of services is desirable,
there are concerns that service jobs are inferior to manufacturing jobs with respect to
compensation and future advancement. These developments have prompted policy-makers to
further explore the nature of these jobs, determinants of service sector growth, and their
potential to generate overall employment growth.
The research literature suggests several aspects of the service industry that should be
examined in order to gain a better understanding of the factors underlying past growth in
services and the potential for future growth. This chapter concentrates on six areas. After a
brief look at employment trends and a short review of some of the relevant literature, we first
137
The job creation potential of the service sector in Europe
explore the shift in composition of occupations within and between services and
manufacturing (Section 1.4). This analysis provides insight into the changing nature of the
two industries in terms of the functions that workers perform and also the extent to which
manufacturing has outsourced service related functions. One of the primary issues related to
outsourcing is the use of temporary help and other flexible workforce arrangements. We
discuss the relationship between the growth in business services and this change in working
relations. Secondly, (in Section 1.5) we examine the sources of demand for services as
revealed through the interaction of manufacturing and service firms, as well as the final
demand of non-business consumers. In addition, we explore the determinants of service
growth by looking at productivity and employment dynamics. Thirdly, we examine (in
Section 1.6) service sector expansion by considering the variation in the size and growth of
the service sector across United States metropolitan areas. Fourthly, we examine (in Section
1.7) the structure of wages within and between services and manufacturing. Fifthly, we briefly
explore (in Section 1.8) the growth potential of services for the economy. The concluding
section summarises our findings and offers policy recommendations.
1.2. Employment growth trends in the service sector in the European
Union and the United States
1.2.1. Comparison analysis
Service sector employment has grown rapidly in both the EU and the United States during the
last decade or so. In both regions, it has outpaced goods-producing employment by a
considerable margin. Between 1985 and 1997, service sector growth in the United States
increased 28 %, while industry employment, which includes manufacturing, construction, and
mining, increased only 3.3 % (Table 1).
Table 1. Employment change, 1985–87
European Community
Total employment
Agriculture
Industry
Mining
Manufacturing
Energy
Construction
Total services employment
Distribution and hotels
Transportation
Finance and business
services
Communal services
1985
1997
133 890
11 195
45 899
1 119
33 361
1 343
10 075
76 797
23 733
8 060
148 820
7 626
44 769
738
31 486
1 230
11 315
96 426
27 550
9 101
9 180
35 824
15 497
44 277
%
United States
1985
1997
%
11.2
– 31.9
– 2.5
– 34.1
– 5.6
– 8.4
12.3
25.6
16.1
12.9
107 197
3 335
30 015
953
20 804
1 270
6 988
73 847
23 822
5 876
129 405
3 502
30 994
700
20 838
1 226
8 230
94 909
28 192
7 179
20.7
5.0
3.3
– 26.5
0.2
– 3.5
17.8
28.5
18.3
22.2
68.8
23.6
10 958
33 191
14 709
44 828
34.2
35.1
Source: Calculations by authors Employment rates report 1998. Employment performance in the
Member States’ European Commission (COM (1998) 572 final).
138
The nature and determinants of service sector growth in the United States
From 1985 to 1997, employment growth in the EU lagged behind the United States. Total
employment in the 15 member countries of the EU (EU-15) increased by only 11.2 % during
the 12-year period, while United States employment rose by 20.7 %. Goods-producing
(industry) employment in the EU-15 declined during the period by 2.5 % while it increased
3.3 % in the United States. The growth trends for service-producing employment in the two
regions are similar. Service-producing employment in the EU-15 increased by 25.6 % during
the 12-year period, compared with an increase of 28.5 % in the United States.
Although much of our analysis will focus on the period from 1985 to 1997, it is useful to gain
a longer-term perspective on employment growth in services and manufacturing in the EU and
United States. Statistics from the OECD go back to at least 1974. The OECD defines services
as the major divisions 6,7,8,9, and 0 of the international standard industrial classification
(ISIC), which include wholesale and retail trade, financial services, transportation, real estate,
public administration, education, health and social work, other community, social and
personal service activities, and private household employment. The industrial (or goodsproducing) sector is defined as ISIC major divisions 2, 3, 4, and 5, which include mining,
manufacturing, construction, electricity, gas and water supply.
Figure 1. Employment trends in the United States and the EU
Employment trends over the longer time period are consistent with the more recent history of
service employment in the two regions. As can be seen from Figure 1, United States service
employment outpaced EU service employment during this time period, increasing 73 %
versus 54 %. The United States total employment grew 46 % while EU total employment grew
only 11 %. Obviously, the slower job growth in services accounts for some of this decline, but
it is not the only reason. Industrial employment in the EU lost jobs during this period,
declining 16 %. Industrial employment in the United States, on the other hand, grew 10 %.
Compounding the problem is the EU’s large industrial base. As shown in Figure 2, industrial
139
The job creation potential of the service sector in Europe
employment in the EU comprised 30 % of total employment in 1997, whereas industrial
employment in the United States accounted for only 24 % of the total.
Figure 2: Employment shares of services and industrial sectors in the EU and United States
During this period, the share of service employment between the two regions converged. The
EU started in 1974 with services comprising 48 % of total employment, compared with a
68 % ratio for the United States. By 1997, the shares for the EU and United States were 65 %
and 73 %, respectively. Thus, during this time, the ratio of EU shares to United States shares
increased from 0.75 to 0.88, clearly marking a narrowing of the gap in service share between
the two regions. On the other hand, the gap between the EU and the United States in
manufacturing employment shares has remained unchanged throughout this period, although
both shares have declined somewhat. In short, differences in employment composition and
not in the variation in growth rates among sectors explains much of the disparity between the
two regions’ total employment growth rates.
1.2.2. Employment in the United States service-producing sectors
The service-producing sectors in the United States, as defined by the standard industrial
classification (SIC), include a broad mix of products and activities. Nearly 80 % of all payroll
employees are engaged in the following six broad service-producing divisions: (1)
transportation and public utilities, (2) wholesale trade, (3) retail trade, (4) finance, insurance,
and real estate, (5) services, and (6) government (Table 2). Much attention has been given to
the category classified as services due to its phenomenal employment growth over the past
several decades. Since 1958, service employment has grown 480 %. By 1998, services
accounted for 30 % of total United States employment. In contrast, manufacturing accounted
for only 15 % of total employment. The increase in service jobs during the last three decades
dwarfs the growth in the next fastest-growing service-producing sector, retail trade. During
the same period, retail trade employment increased 190 %, which only slightly outpaced the
140
The nature and determinants of service sector growth in the United States
145 percent growth in total employment. Manufacturing employment, on the other hand,
grew by only 17 %.
Table 2: Definitions of service-producing categories under the standard
industrial classification (SIC) system
SIC
40 to 49
50 to 51
52 to 59
60 to 65
60 to 62
63 to 64
65
70 to 89
72
73
80
N.A.
Service-producing industries
Transportation and public utilities
Wholesale trade
Retail trade
Finance, insurance and real estate
Finance
Insurance
Real estate
Services
1.1.1.1.1.1.1.1. Personal services
Laundry, cleaning
Beauty shops
Photo studios
1.1.1.1.1.1.1.2. Business services
Advertising
Credit reporting
Janitorial
Employment services
Computer services
Security services
1.1.1.1.1.1.1.3. Health services
Government
% of United
States workforce
5.3
5.4
17.7
5.9
2.9
1.9
1.2
30.0
0.9
6.9
7.8
15.7
Among the major two-digit industries within the service sector, the fastest growing by far is
business services (SIC 73) (1). As shown in Figure 3, business service employment has grown
nearly 1 600 % since 1958. It has increased from 542 000 workers in 1958 to 9.1 million in
1998.
(1) Business services include advertising, mailing, reproduction and commercial art, services to dwellings and
buildings, equipment rental and leasing, personnel supply services, computer programming, data processing
and other computer-related services, to name the larger components.
141
The job creation potential of the service sector in Europe
Figure 3. Service employment trends, the United States 1958–98
This meteoric growth has raised its share of service employment from 8 % in 1958 to nearly
25 % in 1998, as displayed in Figure 4. As a result, business services have played a major role
in total job creation. From 1958 to the present, business services have generated 12 % of all
net new jobs in the economy, while comprising only 1 % of total employment at the start. In
the more recent years between 1985 to 1998, business services accounted for 17 % of total
employment growth. Once again, its contribution to total net job creation far exceeded its
share of total employment.
In contrast, health services (SIC 80) and personal services (SIC 72), the other two major
service industries as measured by employment share, have not come close to matching the
growth of business services. The health service industry comes the closest, growing 650 %.
The personal service sector, on the other hand, has been relatively lacklustre with a growth of
only 50 % during the last four decades. As a result, health services’ share of total service
employment has edged up slightly from 20 to 25 %, but personal services’ share of total
service employment has declined from 12 to 3 %.
142
The nature and determinants of service sector growth in the United States
Figure 4. Components’ shares of service employment
1.3. The nature of services and its implications for growth
This section surveys the literature on services in order to provide a framework for analysing
the growth of services within the United States and to compare this growth to that in the EU.
We offer several hypotheses of the determinants of growth in the services, and then in
subsequent sections we explore how these factors relate to the United States service industry.
Because of the growth and size of the business service sector in the United States, as described
in the previous section, much of our discussion and analysis will focus on business services (2).
In order to provide a framework for understanding the differences in growth between serviceproducing and goods-producing sectors, it is first necessary to delineate the key features that
differentiate these two broad sectors. Martinelli (1991) summarises much of the thinking
about the nature of services by identifying three interrelated characteristics that distinguish
services from goods (3).
(1) Services possess an immaterial nature. Service activities primarily entail a labour input or
performance, which does not yield any immediate material transformation.
(2) Strictly speaking, business services are defined as SIC 73. It includes the activities listed in footnote 66. In
using this definition of services, we attempt to capture those services that are used by businesses, primarily as
intermediate inputs. Of course other services are purchases and provided internally by businesses. Some
researchers (for example, Drennan, 1991) extend the list to include communications (SIC 48), banking (SIC
60), insurance carriers and agents (SIC 63 and SIC 64), legal services (SIC 81) and other related services. This
expanded list of services purchased by businesses is generally referred to as producer services. It is difficult to
obtain data on all these components of producer services for metropolitan areas and for extended time series,
which form the basis of some of our analysis in this paper. Therefore, in most cases we use the definition of
business services (SIC 73), but we lapse at times into using the term producer services when referring to
services as intermediate inputs by businesses.
(3) Sapir (1991) describes services in a similar way when constructing a theoretical framework for understanding
the role of services in the current economy.
143
The job creation potential of the service sector in Europe
(2) Service transactions generally involve simultaneous production and consumption of
services, which requires a direct interaction between the producer and the consumer. In
many cases, the temporal simultaneity of consumption also requires close proximity of
services to their market.
(3) Consequently, services cannot be stored or shipped on their own but are embodied either
in suppliers (in a potential form) or in consumers, or in goods to which services are inputs.
Although most services share these characteristics, Martinelli’s characterisation may be
somewhat extreme for some services. Services, such as banking and other business-related
services as well as legal and health services, are exportable. As we have observed,
employment in these services has expanded considerably in recent years. Services differ in
other dimensions, as well. The different classifications used by various countries and
organisations group service activities by function and by the end purchaser of services. Most
classifications distinguish between consumer services and producer services. In addition,
some classifications, such as the general industrial classification of economic activities within
the European Communities (NACE), combine all services necessary to ensure the general
functioning of the socioeconomic system, such as police service, into a separate category
labelled infrastructure services. These services are consumed by both consumers and
producers. Within producer services, the NACE also distinguishes between services that are
clearly directed to other firms and services that ensure the circulation of capital (financial
sector), goods (wholesale intermediation and distribution), information
(telecommunications), and people (transportation) (4).
Business services include activities that are clearly directed to other firms, whether they are in
agriculture, manufacturing, or even services. These activities include many functions that a
firm could provide internally. However, there are sufficient economies of scale and scope to
warrant the establishment of separate firms to provide these services to the broader market,
not just within their own organisations. The financial services industry and the
telecommunications industry also provide services to businesses. Some classifications, such
as the NACE, aggregate these services into a sector referred to as distributive services.
However, the SIC, which was used by the United States throughout the period studied, does
not classify them under business services. One feature that distinguishes distributive services
from other business services is the typically large investment in fixed capital by distributive
service firms compared with business service firms. Recent consolidations in the financial and
telecommunications field in the United States suggest that there are strong economies of scale
in these industries. Large capital requirements result in economies of scale and consequently
many industries included in this category are regulated. Business services, such as computer
(4) We do not attempt to reconcile the differences in the various classifications used by different countries. To
compare trends, we use the numbers prepared by the OECD, as shown previously in section II. Since we
concentrate on the historical growth of services in the United States, we follow the SIC for most of the other
analysis presented in this study (Figure 1). Although the United States has changed to the NAICS, only
historical data based upon the SIC are available. The SIC lists three major categories of services: personal,
medical, and business. As mentioned in the first section, we will focus on the business services sector, because
of its phenomenal growth.
144
The nature and determinants of service sector growth in the United States
services and professional services, remain fairly small due to their comparative advantage in
conducting unique, specialised activities. There are exceptions, however. Software
development, classified as a business service, can achieve a large scale of operation, as
evidenced by Microsoft Corporation.
Based upon this classification, it is apparent that growth in producer services is closely tied to
the development of the production sector and its demand for non-material inputs. Growth may
also depend upon the industrial structure of the service industry itself. The NACE
classification highlights differences in scale between various service industries, which in turn
affect the ability of firms to enter these industries. Entering an industry, such as
telecommunications, which traditionally requires large fixed costs will be more difficult than
entering an industry with relatively small initial investments, such as a cleaning service. In
addition, growth may also be influenced by political factors in service industries like
telecommunications and transportation that are characterised by economies of scale due to
large fixed costs. These industries are typically subject to government regulation. Excessive
regulation may impede the growth of these service industries. For instance, the
telecommunications industry in the United States grew considerably after the break-up of AT
and T in 1984. Although the industry has recently moved toward greater consolidation, many
experts believe that many advances in telecommunications, including cellular phones and
even the Internet, would not have been as successful if the industry had not been deregulated.
The same can be said of transportation, particularly airlines and trucking.
Therefore, in order to understand the forces driving the fastest growing service sector industry,
it is important to recognise the changes in the industrial organisation of the productive sector.
Martinelli (1991) posits four major developments in industrial structure that have led to the
growth in producer services: (1) the progressive concentration of capital, (2) the rise of the
modern large multi-product, multinational corporation, (3) the growing internationalisation of
markets and competition, and (4) the development of information technology. These
developments have impacted the growth in producer services primarily by accelerating the
trend toward greater division and specialisation of labour. Stigler’s (1951) well-known
theorem on division of labour predicts a transition from performing speciality functions within
an organisation to supplying them in the market. As the market for a given internal phase of
the production process reaches a sufficient size, independent suppliers will find it profitable
to enter the market to provide these services. Consequently, a rise in the service sector may
be the result of services that were once internal to the firm now being provided by independent
suppliers.
The need for large corporations to control, coordinate and monitor their internal operations has
brought about an increase in the demand for managerial, professional, and clerical workers.
In addition, large corporations require information systems that can tie together their far-flung
operations. Thus there is an increasing demand for information systems personnel who can
manage the flow of information. The internationalisation of business has increased the
demand for telecommunications and information processing. International competition has
also increased the need for research and development to improve productivity and for
marketing to promote brand recognition and loyalty.
145
The job creation potential of the service sector in Europe
Martinelli rightfully points out that growth in these activities has occurred both inside and
outside the firm. Therefore, in an attempt to better understand the forces driving the increase
in service activities, it is important to examine the change in the structure of occupations
within the traditional production sectors of manufacturing and within the business service
sectors. We will examine in Section IV the shift in occupational composition within and
across the manufacturing and service sectors to see the extent to which functions performed
within manufacturing firms are now being provided by service industries. We will also
examine United States national input-output tables to determine the sectors that are
purchasing and selling services.
As independent suppliers increasingly find market opportunities for providing service
functions, the next question is where do these firms locate and what factors drive location
decisions. The characteristics of services, discussed so far, suggest that services should be
locally produced and that they should be located close to the centres of production. As
summarised by Martinelli, services are perishable in the sense that they cannot be stored nor
shipped. Furthermore, production and consumption of services are simultaneous. Therefore,
based on the characterisation of services as perishable activities, services would most likely
be produced in close proximity to their customer base. If the primary engine of growth of
business services is demand from industrial sectors, then business services will locate close to
production centres. While business services may have evolved from their close association
with production, the extent to which business services rely on the manufacturing sector
remains to be seen. As the organisational structure of the service sector alone becomes more
complex and intertwined with the rest of the economy, the demand for business services may
come from other sources. We will see in the next section that only 20 % of the demand for
business services comes from the manufacturing sector.
Service activities may be attracted to large metropolitan areas, or smaller areas with excellent
air or surface transportation service. On the other end of the scale, for services that benefit
from large economics of scale, such as financial processing activities (known also as
backroom activities) or telemarketing, cost considerations are a factor and more rural
environments can be attractive.
In addition, most researchers view the growth of international trade and the growth of
multinational corporations as pushing service activities to larger cities in developed
economies. Many services play complementary roles to the headquarters of multinational
corporations and are increasingly playing a part in the site selection for advanced production,
research and development facilities and for headquarters (5). Moreover, while advancements
in new information technology provide the potential for the de-concentration of producer
services, they can equally promote further concentration. As long as the urban and
(5) Drennan (1992), in an analysis of metropolitan sources of growth in the export of producer services, emphasises
the importance of agglomeration economies. He finds that the largest producer service firms are headquartered
in a country’s largest nodal cities, which indicates the existence of strong agglomeration economies. Thus, a
country’s international competitiveness in producer services is linked to the economic and social viability of a
nation’s largest cities.
146
The nature and determinants of service sector growth in the United States
agglomeration economies of metropolitan areas remain strong, improvement in information
technologies may only enhance the current trend toward concentration.
Martinelli argues that economic trends will only reinforce the tendency for producer services
to locate in a handful of large metropolitan areas in developed countries. In fact he makes the
point that producer services are more concentrated in these areas than is manufacturing, and
highlights the two standard arguments used to explain these trends: (1) the hierarchical
structure of large corporate organisations and (2) the peculiar agglomeration economies linked
to service transactions. Moreover, the existence of robust business and producer service
sectors generates additional benefits including increased productivity within the entire
productive system through specialisation and agglomerations and increased growth in
employment and human capital. The potential for producer services to affect the productivity
of the businesses they serve, principally manufacturers, once again underscores the linkage
between service-producing and goods-producing sectors. This linkage has implications for
measuring productivity of the two sectors. Instead of considering the productivity of the two
sectors separately, as is typically done, one should consider the effect of business services on
the productivity of the goods-producing sector as a measure of the productivity of generating
business services (Kuznets, 1971 and Porter, 1990).
Jaeger and Durrenberger disagree with Martinelli and others in suggesting that economic
trends and corporate restructuring are all pushing toward increased concentration. The
authors found a more complex arrangement where multiple hierarchies of city patterns exist
where cities concentrate in specific industries. Their research of central Europe found counter
urbanisation forces that foster situations of multiple hierarchies similar to that of the United
States. Still, the authors concluded that the outsourcing of service activities from a declining
manufacturing base tends toward ‘concentralisation’.
1.4. Structure of employment growth: Shifts in occupational demand
As discussed in the previous section, research points to structural changes in the business
sector as one of several factors responsible for service sector growth. The changing industrial
structure of the business sector can be detected through changes in the occupational
composition of its work force. Since workers’ occupations are classified by the functions they
perform, it is possible to trace functional activities through occupation composition. For
example, a stagnant industry that is undergoing little technical or managerial change will
maintain its occupational composition through time. It is likely that its employment base will
slip as it fails to adjust to changing markets, but its occupational composition will hold steady.
In sharp contrast, a more dynamic sector, which is taking on new markets and adopting new
technologies, will most likely have a rapidly changing occupational mix (6).
(6) It is important to note that occupational definitions may also change, which would dampen the ability to detect
structural change through examining an industry’s changing occupational mix. For example, clerical workers
can become skilled in computer processing and data entry but maintain their clerical job titles.
147
The job creation potential of the service sector in Europe
To measure compositional change of occupations, we use a modification of the Lawrence
Index of industrial change. The measure is the absolute difference in the share of each
occupation between 1984 and 1997, summed over all occupational categories. A value of zero
indicates no change in the occupational composition of the industry, and a maximum value of
two represents a 100 % change in occupational mix. As shown in Table 3, the United States
business services sector witnessed a relatively large change in its occupational composition
from 1984 to 1997. At the same time, the occupational composition of manufacturing was
much more stable, in that its change was no more than the change for all occupations and all
industries within the economy.
Table 3. Index of the change in occupational composition, 1984–97
Sector
Total
Business services
Manufacturing
Index (1)
0.196
0.392
0.189
(1) Index is based on the Lawrence index, in which a value of 0.0 shows that the occupational
composition of the sector did not change between the two years, while the maximum value of 2.0
would indicate a 100 % shift in the sector’s occupational composition.
1.4.1. Occupational shift-share analysis
A more detailed shift-share analysis can provide more information regarding changes in a
sector’s industrial structure and its changing occupational mix. Shift-share analysis separates
an industrial sector’s growth in occupations into the following three components (7):
National growth — this component reflects overall growth in the domestic economy. It is
computed assuming that each occupation grows at the national growth rate of total
employment. A growing economy will support a certain degree of employment growth across
all occupations in all industry sectors.
Change in occupational mix — industries with a positive occupational mix employ a large
number of individuals in occupations that are experiencing faster than average growth
nationwide. This component is computed by multiplying each occupation by the difference in
the national growth rate of that occupation and the national growth rate of total employment.
There are several explanations why an industry’s occupations may grow faster than the
national average. First, the industry may be an intensive user of occupations that are in strong
demand across many sectors. For example, an industry that has been a long time user of
computer analysts will have a positive occupational mix as more and more industries hire
computer analysts. Second, the industry may be using a large number of occupations that by
(7) Shift-share analysis is typically used to examine changes in industrial composition across geographical areas.
We use it here to look at the change in occupational mix within industries.
148
The nature and determinants of service sector growth in the United States
their nature have low productivity levels but are in high demand. For example, food service
workers are a fast-growing occupation, in part, because of their low productivity; therefore an
industry that hires a large number of these workers will have a positive occupational mix.
Change in the occupational share — this component measures the difference between the
national rate of growth of an occupation and its rate of growth within a specific industry.
Industries with a positive occupational share component experience above average growth in
their major occupations. Due to their competitive advantage, these industries provide a more
conducive environment for these occupations than exists elsewhere. For example, while
computer occupations are in high demand across many industries, their growth has excelled
within highly specialised firms in the business services industry. This occurs, in part, because
many firms in other industries are outsourcing their computer needs to these specialised firms.
In Table 4 (located in the appendix at the end of the chapter), the change in occupational
employment in the United States from 1984 to 1997 is broken down into the above three
components. For each of the 14 industries shown in Table 4, employment trends for 50
occupations have been calculated. Turning to both durable and non-durable goods production,
the analysis suggests that approximately a third of the gross employment loss in
manufacturing is due to shifts in occupational demand as shown in its occupational mix
component. Changing business conditions are calling for growth in specific occupations that
are not highly concentrated in goods-producing sectors. Obviously, manufacturers are less
likely to provide legal, health, or retail services as part of their core activities; hence, it is not
surprising that they are losing ground in the faster-growing occupations.
This analysis also captures the movement of seemingly blue-collar production workers from
manufacturing to business and professional services through the use of temporary and leasing
employment companies. Note that among the top twenty occupations lost by manufacturers,
several, including fabricators and assemblers, freight stock and handlers are production
occupations. At the same time, as shown on Tables 5, 6 and 7, (located in the appendix at the
end of the chapter) fabricators and assemblers are among the top twenty occupations for which
business service firms have gained share. Other production level occupations ranked among
the top twenty within business and professional services include labourers excluding
construction (8), and machine operators and or tenders. In short, this analysis supports
researcher’s claims that in the United States a share of the growth in services is coming from
the outsourcing or externalising of activities to the service sector. Moreover, the analysis also
points to the growing importance of temporary employment services in staffing productionrelated activities in manufacturing firms.
(8) It is interesting to note that temporary employment firms, in general, avoid contracting with construction firms
due to work injuries and liabilitys issues.
149
The job creation potential of the service sector in Europe
1.4.2. The outsourcing of non-core support services in the manufacturing sector
The above shift-share analysis suggests that many manufacturers, facing highly competitive
environments due to a global overcapacity in many core industries including steel,
automotive, paper, and semiconductor, have outsourced their non-core service activities to
independent providers. In addition, the growth of speciality producer service providers is
pulling more service functions out from under the manufacturing umbrella.
Monnoyer and Philippe (1991) argue that much of the growth in producer services is due to
large manufacturing organisations externalising their demands for services. Originally
pursued as a cost-saving strategy, the practice grew as manufacturers found that specialised
service companies offered more expertise in providing the needed service than could be
supported in-house. Monnoyer and Philippe (1991) further suggest that structural factors may
contribute to the provision of services by small establishments in large headquarter cities.
Moreover, there is particular evidence suggesting that manufacturers are using more and more
temporary employment services in addition to more flexible in-housing staffing arrangements.
1.4.3. The use of temporary services
A very small percentage of the United States’ employed workers work for temporary
employment agencies. Yet, it is one of the nation’s fastest growing industries. In 1999,
temporary employment agencies employed 3 million workers or 2.4 % of the nation’s
workforce, up from only 1 million workers or 0.9 % of the employed work force in 1990.
Temporary employment agency workers, as well as leased workers, contractual workers and
self-employed workers, are in work arrangements where the workers’ on record employers are
different from the organisations and or firms where they work. This, in large part, explains
how business services and professional services can house a growing share of productionrelated occupations as shown above in Tables 6 and 7.
In their national survey of employers, Erickcek and Houseman (1997) find that a large
percentage of the surveyed manufacturers, 72 %, used temporary employment agency
workers. Table 8 provides the major reasons given by all surveyed employers regarding the
use of temporary help agencies. More than 50 % of the surveyed firms cited the need to meet
unexpected demand as one of the reasons for using temporary employment agencies. Indeed,
more and more manufacturers are finding that their orders are becoming more variable, while
intense competition is pushing them to better match their staffing levels to output activity.
Many of the employers surveyed use temporary agency workers for the more traditional
reason of filling in for absent permanent workers. Finally, while not among the top five
reasons, 21.3 % of the employers surveyed use temporary employment services as a screen for
potential permanent hires.
In addition to the cost saving generated by being able to more closely match their employee
levels to the changing volume of orders, employers also save on hourly labour costs by using
temporary employment agency workers. Hourly wage costs of using temporary employment
agencies workers are not significantly different from hiring a permanent worker, when the
150
The nature and determinants of service sector growth in the United States
agency service fee is factored in. However, large savings can occur because the employer
does not incur benefit costs for the temporary employment workers. In fact, Houseman (1996)
finds that the use of temporary employment services was positively related to the company’s
level of employee compensation for its permanent workers; better-paying companies are more
likely to use temporary employment workers than lower paying firms.
Table 8. Major reasons for using temporary help workers
Reasons
Provide needed assistance at times of unexpected increases in business
Fill in for absent regular employee who is sick or on vacation
Fill vacancy until regular employee is hired
Special projects
Seasonal needs
1.1.1.1.2 %
responding
52.2
47.0
46.6
36.0
28.1
Source: Erickcek and Houseman (1996).
The experience of European nations parallels that of the United States. Bronstein (1991) also
finds that workers under temporary work relationships accounted for a very small portion of
the overall workforce in western Europe. In addition, through his interviews of employers and
temporary employment agencies, he found that employers in western Europe used temporary
employment for the same reasons as employers in the United States.
In a recent study of German employment relations, Hoffman and Walwei (1999) also found
growth in flexible staffing arrangements, although such arrangements represent a small
percentage of German workers. Hoffmann and Walwei point out that the growth has been due
to the increase in part-time workers and self-employed individuals and not to employers
increasing their use of temporary, leased or contractual workers. In addition, the authors could
find only a modest relationship between the growth of non-regular work arrangements and
the changing industrial and gender structure of the German economy. The authors interpret
this to mean ‘that even irrespective of the sector-specific and gender-specific changes in the
employment structure, regular employment relationships, part-time work and forms of selfemployment would have developed in the same direction and to roughly the same extent.
Instead the authors point to: (1) more relaxed labour regulations which allow for greater use
of flexible staffing arrangements, (2) the greater use of these arrangements to save on
adjustment costs of meeting fluctuations in production, and (3) the need to lower unit
compensation costs by outsourcing non-core activities.
Hansen (1990) disagrees that outsourcing can explain the rapid growth of producer services.
Instead, he turns to earlier research that suggests that a substantial share of producer service
growth can be attributed to changing business practices, which in turn are changing the
structure of production. Manufacturers are not outsourcing old in-house activities. Rather,
they are using new services to meet new challenges that require skills and expertise that were
never developed in-house. Although the presence of producer service firms provides potential
151
The job creation potential of the service sector in Europe
cost-reducing opportunities for other firms, it would probably be more accurate to argue that
it is typically the changing demands of firms that provide markets for specialised service
functions, with the whole process increasing the division of labour.
1.5. Sources of service employment growth
While important in explaining the lacklustre growth in manufacturing, structural change in
manufacturing appears to have only a modest impact on the growth in services in the United
States. As shown in Table 9 below, services continue to account for a small, although growing,
share of manufacturers’ intermediate, inter-industry purchases. The selected services share of
all intermediate goods purchased by manufacturers stood at 8.4 % in 1992 compared to 7.4 %
in 1982. Business services, which include temporary employment agencies, and janitorial and
security services witnessed the largest gain in share from 5.2 % in 1982 to 6.4 % in 1992.
Table 9. Share of manufacturers’ intermediate demand for goods and services
Industry
1982 (%)
1992 (%)
Communications
Finance and insurance
Business services
Health, education, and social services
0.8
1.2
5.2
0.2
0.4
1.4
6.4
0.2
Intermediate inputs
100
100
Source: 1982 and 1992 — input-output tables, United States Bureau of Economic Analysis.
To explain the robust growth in service-producing sectors both in the European Community
and in the United States, researchers have focused on several key factors: (1) internal growth
in services, (2) the high income elasticity of demand for consumer services, (3) the lack of
productivity growth in services, and (4) employment dynamics in explaining the robust
growth in service-producing sectors.
1.5.1. Internal growth in services
The growth in services creates increased demand for services, through the purchase of
services by service-producing businesses. As shown in Table 10, non-manufacturing activities
account for 65.8 % of total demand for business services in 1992, up from 59 % in 1982. While
a larger percentage of manufacturers’ intermediate purchases are for business services (Table
9), their share of total service receipts fell during the ten-year period. For instance,
manufacturers accounted for more than 20 % of total business services sales in 1982. In 1992,
the share dropped to less than 14 %. Similar reductions in share are reported in
communications and finance and insurance. Finally, personal consumption is the major source
of demand for health, education and social services and is a major market for finance and
insurance and communication. Beyers (1991) reports similar findings in his 1989 study:
152
The nature and determinants of service sector growth in the United States
‘… manufacturing is not an important market for most producer services. In fact, household
consumers are very important customers for financial, insurance, real estate, and legal
services; government is a very important client for accounting, labour supply, research and
development, and consulting services. The various services are more important inter-industry
markets from most producer services than are the goods producing sectors.’
We will further explore the relation between manufacturing and service growth in our
subsequent analysis of service sector expansion in metropolitan areas.
Table 10. Composition of demand for services by type, percentage of total demand, 1982–92 (9)
Manufacturing
NonPersonal
%
manufacturing consumption
1982
Communications
Finance and insurance
Business services
Health, education, and
social services
1992
Communications
Finance and insurance
Business services
Health, education, and
social services
Government
%
11.2
6.3
20.5
35.7
40.4
59.0
41.1
48.2
8.4
5.7
3.0
9.3
0.7
3.2
93.9
2.0
3.7
4.1
13.9
42.4
36.7
65.8
43.9
54.0
8.3
5.7
2.4
8.5
0.5
3.2
105.7
– 9.4
Source: 1982 and 1992 — input-output tables; United States Bureau of Economic Analysis.
1.5.2. The high income elasticity of demand for consumer services
Although the primary focus in this chapter is on business or producer services, it is useful to
digress momentarily to look at consumer services. As shown in Table 10, consumer demand
accounts for a large share of service output, except for business services. In addition,
consumer demand for services, such as recreation, health, and personal and or household
services, increases with disposable income. As shown in Table 11, the GDP per capita from
1988 to 1997 was more robust in Europe than in the United States. Growth in GDP per capita
in the United States increased by 50 % during the period. Only Ireland, Finland and Sweden
were unable to surpass the United States during the period. Therefore, it is not surprising that
employment growth in the community, social and personal services as well as in distribution,
(9) This table provides the percentage distribution of total demand by type of user for each of the services listed in
the first column. The percentages across each of the rows do not add up to 100 % because several smaller
demand categories are excluded, for example gross private investment and net exports. Second, because State
and local governments are major producers and not users of health and education services, their demand for
these activities is negative. This, in turn, forces personal consumption demand for health, education and social
services to be recorded as being greater than 100 % of the total output.
153
The job creation potential of the service sector in Europe
retail and recreation are similar in Europe and in the United States. Employment in hotel,
retail and recreation grew 16.1 % in the EU-15 compared to 18.3 % in the United States.
Employment in community, social and personal services rose by 23.6 % in the EU-15, which
was below the 35.1 % increase in the United States.
In the United States, one factor affecting the growth of consumer-originated service is the
continuous expansion of the urban fringes of many metropolitan areas. As retailers follow
their customers and strive to meet their shopping needs, they continue to build in more rural
reaches of the urban fringe and stay open for longer hours. The combination of longer store
hours and the continuous pursuit of customers to the outer fringe of expanding metropolitan
areas may generate strong retail employment growth in the future that cannot be explained by
income growth alone.
Table 11. Percentage change in per capita income, 1988–97, in USD and %
1988
1997
1.1.1.1.2.1.1.1
% change
Belgium
Denmark
Germany
Greece
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
Finland
Sweden
United Kingdom
12 623
13 682
14 161
6 799
9 343
13 603
16 117
12 985
15 558
12 832
12 506
6 750
13 798
14 772
13 428
23 242
25 514
22 049
13 912
15 990
21 293
20 634
21 265
33 119
22 142
23 077
14 562
20 488
20 439
20 483
84.1
86.5
55.7
104.6
71.1
56.5
28.0
63.8
112.9
72.6
84.5
115.7
48.5
38.4
52.5
United States
19 558
29 326
49.9
Source: Historical statistics, 1960–98; OECD Economic Outlook, 1990.
1.5.3. The lack of productivity growth in services
Although hindered by serious measurement problems, many researchers argue that
employment growth in services, especially employment growth in relatively low-wage jobs,
is due to the lack of productivity gains in services.
154
The nature and determinants of service sector growth in the United States
Figure 5. Productivity trends in United States manufacturing and services, 1980–96
As shown in Figure 5, productivity in United States services, as measured by real output per
worker, has modestly declined during the past 18 years. In contrast, manufacturing
productivity continued to increase during the period. In percentage terms, service
productivity declined by 14 % in the 18 years, while manufacturing productivity rose by a
healthy 81 %. With the demand for nearly all services growing, it seems clear that the overall
decline in productivity has only heightened the demand for more service workers. At the same
time, manufacturing productivity gains have outpaced output increases and have placed
downward pressure on the need to hire additional workers. Thus, the productivity differences
between services and manufacturing alone account for a significant portion of the gap in the
growth of service and manufacturing employment.
The contribution of productivity trends to the growth in services can be illustrated by asking
the hypothetical question: What would be the growth rate of service employment if
productivity in the services grew at the same rate as productivity in manufacturing? This
comparison has some relevance, since in 1980 output in the service sector (narrowly defined)
was about the same magnitude as output in the manufacturing sector. Furthermore, output of
the two sectors grew at about the same rate throughout the 17-year period; a 72 % increase for
services compared to a 66 % increase for manufacturing. Obviously, if service sector
productivity approached manufacturing productivity, employment growth would be
comparable. These actual and hypothetical growth trends are illustrated in Figure 6.
One can see that the hypothetical growth rate in service employment (labelled hyposrv) would
be about the same as the actual growth in manufacturing employment (labelled actman) if the
two shared the same productivity trend. The converse is also shown in the figure.
Manufacturing employment growth would grow at roughly the same rate as service sector
growth if manufacturing productivity were as sluggish as service sector productivity. Of
course, if manufacturing productivity were to follow the same trend as service sector
productivity, the increase in manufacturing output would probably not be as great because of
155
The job creation potential of the service sector in Europe
greater cost pressures and less competitiveness. If the service sector shared manufacturing’s
healthy productivity increase, its output would have grown even faster.
Figure 6. Actual and hypothetical employment productivity trends, 1981–95
1.5.4. Employment dynamics
Job creation in the service sector accounts for a significant share of the new jobs in the
economy. The generation of these new jobs comes about from the birth of new establishments
and the expansion of existing ones. In order to better understand the factors that contribute to
the creation of new jobs, it is instructive to examine the dynamics of employment change.
Examining gross flows provides insight into the relative contribution of births and expansions
to service job creation and the relative effects on closings and contractions to service job
destruction. Such information may be useful in fashioning policy to attract and retain
businesses.
Calculating gross employment flows requires tracking individual establishments over time in
order to observe changes in their employment status. Individual establishment-level census
data are confidential and can be analysed only at specified census offices. We contracted with
the census to prepare special tabulations of gross employment flows at the State and
metropolitan level for one-digit and two-digit industries. At the time of the request, only a few
years of data were available and the most recent year was 1992. As 1992 was a recession year,
we chose the 1989–90 period. Net employment change within each industry over the time
period is divided into four components: (1) employment gains due to the opening of
businesses, (2) employment gains due to the expansion of existing businesses, (3)
employment losses due to the closing of existing businesses, and (4) employment losses due
to the contraction of existing businesses (10).
(10) Employment due to openings is defined as the employment level in 1990 of an establishment that did not exist
in 1989. The definition of a closing is the opposite; employment in 1989 of an establishment that ceases to
exist in 1990. Employment due to expansions is calculated as the difference in employment of establishments
that expand between 1989 and 1990. Contractions are calculated the same way.
156
The nature and determinants of service sector growth in the United States
Table 12 displays the components of net employment change for manufacturing, services
(narrowly defined) and business services (SIC 73). Several features about the employment
dynamics of services stand out in comparison with the manufacturing sectors. First, the
service sector is more volatile than the manufacturing. Summing the absolute value of the
percentage change in employment of all four components yields a turnover rate for each
sector. This turnover rate measures the percentage of jobs that changed hands between 1989
and 1990. For the manufacturing sector, 22 % of jobs turned over within the one-year time
span. For services, the turnover rate was 31 % — 9 percentage points higher. For business
services, the turnover rate was more than double that of manufacturing, posting a 49 % rate.
The high volatility of the service sector is also apparent in the sheer number of jobs lost.
Between 1989 and 1990, the service sector added 1.3 million net jobs to an economy that
gained a net total of 1.8 million jobs. Yet, it lost nearly three times as many jobs due to closings
and contractions as it created on net. Manufacturing lost an even higher percentage, but this
is to be expected given the net loss in manufacturing jobs.
Table 12. Net employment change by industry sector in the United States, 1989–90
Openings
Change
% change
Closings
Change
% change
Expansions
Change
% change
Contractions
Change
% change
Net
Change
% change
Turnover rate
% change
Total employment, 1989
Manufacturing
Services
Business services
401 403
2.05
– 671 358
– 3.45
1 581 313
8.09
– 1 684 951
– 8.62
– 376 593
– 1.9
22.21
19 545 070
863 818
3.1
– 1 245 120
– 4.51
4 118 916
14.93
– 2 441 625
– 8.85
1 285 989
4.7
31.39
27 585 222
266 142
5.4
– 352 233
– 7.15
1 094 825
22.22
– 681 944
– 13.84
326 790
6.63
48.61
4 928 143
NB: Turnover rate is the estimation of its rates of change of the four components of employment
change.
Source: Bureau of the Census, Census-based small business database, special tabulations for the
Upjohn Institute, 1989–90.
For all three industries displayed in the table, expansion of existing businesses is the primary
source of job creation. The preponderance of expansions is to be expected, as the time period
spans only one year. One year leaves little time for new firms to start up. A longer time period
would shift the relative contribution toward openings. However, it is still interesting to
examine the difference, or in this case, the similarity in the characteristics of job creation
across sectors. Business openings account for roughly 20 % of jobs created. The
manufacturing sector experiences a slightly higher percentage, but about the same as business
services. The greater role of business openings in job creation in manufacturing than services
157
The job creation potential of the service sector in Europe
is somewhat surprising since manufacturing start-ups typically require more capital
investment. Manufacturing firms are significantly larger than service firms, with 57
employees per firm compared with 15 employees per firm in services. Even business services,
which include firms like Microsoft Corp, have only slightly more employees per firm,
approximately 19.
1.6. Empirical evidence of the manufacturing and/or services nexus
1.6.1. Geographic characteristics of service employment growth
As noted in the literature review in Section II of the chapter, location of service activity should
be tied to several factors: (1) proximity to customers, including production activities, (2)
sufficient market size to spawn independent suppliers, (3) favourable changes in the
organisation of an industry that increase demand for services, and (4) agglomeration
economies. We will examine these factors by focusing on 78 of the largest metropolitan areas
(MSAs) in the United States. These areas comprise about 43 % of United States employment
and 50 % of United States employment in services. These MSAs also account for the same
proportion of total growth and service growth from 1985 to 1997. A list of these metropolitan
areas is included in the appendix (11).
To estimate the determinants of employment growth in the service sector, we first consider the
change in the share of service employment across the 78 MSAs between 1985 and 1997.
Service employment as a share of total employment increased from an average of 27 % in
1985 to 35 % in 1997. These percentages are higher than the national average of 23 % in 1985
and 29 % in 1997, which is to be expected if services benefit from large markets and
agglomeration economies. The service share of total employment ranges across the
metropolitan areas from a low of 24 % in 1997 for the Cleveland metropolitan area to a high
of 51 % for Las Vegas (known for its gaming and entertainment industry). The range was
wider for 1985, with Las Vegas topping the list with the same high percentage of 51 % but
Charlotte, North Carolina occupying the bottom with 18 %. It is interesting to note that the
service sector in Cleveland, while claiming the lowest employment share in 1997, maintained
the same 24 % share throughout the 12-year period. Its decline in the ranking from 46th in
1985 to last in 1997 was affected by a faster increase in service share of employment by the
33 metropolitan areas that were listed below Cleveland in 1985. Therefore, to begin to
understand the factors of growth, we will examine the change in the share of service
employment from 1985 to 1997.
(11) We chose to define regions as metropolitan areas. Another choice, and one often used by researchers doing
regional analysis within the United States, is to use States. States are convenient because data are collected by
States and for the purpose of this paper, their population size corresponds more closely to European countries.
However, most regional economists agree that metropolitan areas approximate a natural labour market more
so than States.
158
The nature and determinants of service sector growth in the United States
Services in all but three of the 78 MSAs expanded their share of total employment between
1985 and 1997. Only Cleveland, Ohio, Little Rock, Arkansas, and Las Vegas registered no
growth in the size of the service sector during this time period. Charlotte, North Carolina and
Augusta Georgia topped the list with increases in excess of 65 %. As illustrated by the map in
Figures 9 and 10 (located at the end of the chapter) most of the metropolitan areas
experiencing significant expansions in the service sector are concentrated in the southern part
of the United States. The States of Texas, Florida, Georgia and the Carolinas dominate. The
number of service jobs in these States’ 15 MSAs doubled from 1.5 million in 1985 to 3.2
million in 1997, with their share of total service jobs in the 78 MSAs increasing from 13 to
17 %.
While service sector expansion has occurred at a faster pace in the south than in most other
parts of the country, one should not overlook the changes taking place within the northern
regions of the United States, which have traditionally been dominated by manufacturing.
Groshen and Robertson (1993) examine the changing industrial structure of 10 metropolitan
areas in the Great Lakes region. MSAs in this region have lost manufacturing jobs, and their
overall employment growth has been sluggish, at least throughout the 1970s and 1980s.
Based on their analysis, the authors come to two interesting conclusions. First, MSAs in the
Great Lakes were not losing factory jobs to other parts of the country. Rather, these cities were
losing factory jobs to smaller municipalities or rural areas in their own region. Several factors
may enter into a manufacturer’s decision to locate outside large cities. They may find the
location advantage of large cities diminishing due to high land costs, lower communication
and transportation costs, shrinking plant sizes, greater awareness of environmental
contamination, and union avoidance.
Second, replacing these manufacturing jobs are service jobs. MSAs in Great Lake States have
increased their share of service sector jobs. While manufacturers find urban areas less
attractive, service providers enjoy the urban advantages of first-class office space, easy access
to information transfer systems, and a pool of seasoned workers from other service firms and
local universities. Furthermore, this suggests that service providers and customers do not have
to exist in close proximity. Manufacturers and other firms are continuing to patronise large
city service providers even from a distance. From this, Groshen and Robertson reach the
general conclusion that services are increasingly exportable, which to them makes sense
considering the technological advances that make more distant relationships possible. They
list several innovations; the introduction of the fax machine and electronic mail, coupled with
a decline in transportation costs.
1.6.2. Factors determining service sector expansion
To explore the factors driving the overall increase in service sector employment share, we
relate the percentage change in service employment share on several factors. These factors
include population, presence of manufacturing activities, establishment size, unionisation, job
creation activity, and characteristics of the workforce. The choice of these variables follows
as closely as possible the factors generally cited in the literature, given available data. We
estimate these relations for both services (narrowly defined as SIC 70–89) and business
159
The job creation potential of the service sector in Europe
services (SIC 73). The results are shown in Table 13 (located in the appendix at the end of the
chapter).
The findings suggest that service sectors expand more rapidly in regions that have a large
manufacturing base, low rate of unionisation, low manufacturing wage rate, and a relatively
educated labour force. As evident in the table, we tried different specifications to see how the
interactions among these variables affect their impact on the service sector. For some
variables, the effects were strong across all models. For instance, service expansion is
positively related to a strong manufacturing base. This relation holds for both services and
business services across all specifications accept one. Even the magnitude of the effects is
stable across models. The importance of manufacturing on service expansion is evident in the
dynamics of employment growth within manufacturing. The business service sector is shown
to expand in those MSAs in which job creation in manufacturing originates predominantly
from openings. This finding is robust across all models. The same variable is also robust
across all specifications for services, but in this case the correlation is negative. It is curious
that the share of job openings in service sector job creation has no statistically significant
relation with the expansion of the service sector.
On the other hand, job creation activity within an MSA, as measured by the percentage of jobs
created through openings and expansions, is statistically significantly related to service and
business service expansion in only one specification. Once union representation and wage
rates are entered, the job creation variable is no longer statistically significant, but it does
retain a positive sign for business services.
Manufacturing wage rates are negatively related to the expansion of both services and
business services within an MSA. This result may reflect the difficulty experienced by service
firms in attracting workers away from high-paying manufacturing jobs. Unions may also
restrict the supply of workers, and this variable is also negatively related to service expansion.
The variable, however, is only statistically significant for services and not for the more
narrowly defined business service industry. Since union representation is lowest in the south,
this variable may be a proxy for MSAs in southern States. Thus, factors associated with
southern MSAs, but not included in the estimation, may be proxied by the union
representation variable.
The education of the MSA labour force is only weakly related to the expansion of the service
sector, with a statistically significant relation between the percentage graduating from college
and the expansion of business services. Other variables, such as an MSA’s unemployment rate
and the employment-to-population ratio, used to measure the availability of labour, were
included in the model, but the results were statistically insignificant and thus were not
included in the table.
1.6.3. Determinants of service sector job creation
We also consider several factors that may be related to job creation in the service sector within
MSAs. We estimate separate equations for the percentage increase in service jobs due to
160
The nature and determinants of service sector growth in the United States
openings and the percentage increase in service jobs due to expansions. This analysis is based
on data for the same 78 MSAs as used in the previous section, but for the shorter time period
of 1989–92.
The results are shown in Table 14 (located at the end of the chapter). The results suggest
several conclusions. First, service sector job creation is closely tied to manufacturing job
creation. In addition, an initially large service sector does not appear to enhance the creation
of service jobs. In fact, in several model specifications, it deters job creation. This result runs
counter to the argument that service companies require a critical mass within local markets.
Nonetheless, larger MSAs witness a greater level of job openings than smaller MSAs,
suggesting the importance of agglomeration economies. The results also suggest that union
representation is negatively correlated with job openings and expansions. This relation is
statistically significant for all specifications. Interestingly, unlike the results in the previous
section on service sector expansion, hourly wage rates are not statistically significantly related
to either openings or expansions.
1.7. Wage structure
While services have accounted for much of the growth in United States employment during
the past several decades, concern has been expressed that these jobs are low-paying and offer
little opportunity for advancement. To address this issue, we examine several dimensions of
the wage structure of workers in service and manufacturing industries. First, we look at the
distribution of wages within the service and manufacturing sectors. Dupuy and Schweitzer
(1994) argue that looking at average wages across industries can be misleading in comparing
wages across industries. Average wage differences are trivial compared to the wide range of
salaries available in each sector. Second, we consider the wages of young workers who
recently entered these industries. Examining wages for the entire workforce within industries,
which combines new entrants and those with years of experience, can mask the opportunities
available to workers who are just starting their careers. Finally, we extend the examination of
entry-level workers to look at the wage distribution over an individual’s entire career. By
comparing age-earnings profiles for a cross-section of workers within manufacturing and
services, we can see whether workers in services have less opportunity for advancement than
workers in manufacturing.
1.7.1. Wage distribution
Dupuy and Schweitzer (1994) examine the wage distribution within services and
manufacturing for 1992. They conclude that the characterisation of service jobs as lowpaying, dead-end jobs is inaccurate. Although they find that the median wage in the serviceproducing sector is about 4 % lower than the median wage in the goods-producing sector, they
also find that a wide range of high-paying jobs is available in the service-producing sectors.
These wages generally match the wages available in the goods-producing sector. Serviceproducing sectors, such as fire insurance, and real estate and public administration, offer
median and top wages that exceed those in manufacturing. The wage differential for young
161
The job creation potential of the service sector in Europe
workers in the service-producing and goods-producing sectors is negligible, differing only by
USD 2 per week. Wages in the narrow definition of services, which includes business services
as well as personal services, are higher than manufacturing wages at the median but not at the
top 90th percentile.
Dupuy and Schweitzer find that the goods-producing sector benefits the less-educated worker.
Workers with only a high school education are paid 20 % higher in the goods-producing sector
than in the service-producing sector. The differential is even higher for workers in services
than for those in manufacturing. The authors conclude that this gap is not an adequate
argument for a policy that would shift the distribution of employment opportunities towards
goods production. Instead, they conclude that these results underscore the importance of an
educated workforce.
We extended their analysis to include two additional time periods, by grouping the years
1984–86 and 1995–97. Three years of data are pooled for each time period in order to
minimise the possibility that the events of one particular year may differ significantly from the
general trend. The weekly wages are presented in 1997 USDs so that wages can be compared
over time. Only workers on company payrolls are included in the sample (12).
The results displayed in Tables 15–20 reveal that median wages in the service-producing
sector are roughly 7 % lower than wages in the goods-producing sector. The top-end serviceproducing jobs are also below the top-end goods producing jobs, but by only 4 %. The wage
differential between the narrow service industries and manufacturing is also 7 %.
Entry level workers (ages 25 to 30) find little difference in median pay between the serviceproducing and goods-producing sectors. The median wages in services are the same as those
in manufacturing, but the top-end service wages are somewhat lower. However, there is a
greater difference at the lower end of the pay scale. Less-educated workers are at a
disadvantage in service producing jobs. The median wage of a service sector less-educated
employee is nearly 20 % lower than the median wage of a similar worker in the goodsproducing sector. For the narrow services, the wage differential is even greater at 28 %, and
the differential for top-end jobs is close to 30 %. Clearly, manufacturing and other sectors
with physically-oriented occupations such as construction and perhaps wholesale trade hold
greater opportunities for the less educated.
The wage gap between service-producing and goods-producing sectors have narrowed over
the decade or so covered in our analysis. Between the mid-1980s and the mid-1990s, the
differential for the median wage in the service-producing sector has declined from 87.6 % of
goods-producing sector wages to 93.1 %. The gap has also closed somewhat in the narrow
(12) Data are obtained from the March current population survey (CPS), which includes yearly information on a
sample of nearly 60 000 households. Wages are measured as weekly earnings of full-time workers. Workers
are considered full-time employees if they work more than 35 hours per week, but not necessarily at one job.
The work week for full-time workers is concentrated at 40 hours, so using weekly earnings or hourly wages
will yield similar results.
162
The nature and determinants of service sector growth in the United States
services industries, both for median wages and for the top-end wages. The same narrowing
has taken place for entry-level workers and for less-educated workers.
Wage growth for all sectors during this time period has been relatively stagnant. The gap has
narrowed because wages in the service-producing sector have not fallen as fast as those in the
goods-producing sector, and in some cases, service sector wages have increased. For all three
groups, inflation-adjusted median wages in the goods-producing sector have fallen during this
ten-year period. The same holds for entry-level workers and less-educated workers in the
service-producing sector. In contrast, wages of all workers in the narrow services industries
have increased overall.
1.7.2. Age-earnings profile
Another concern about service sector jobs is the lack of wage growth over a worker’s career.
One way to examine relative opportunities for advancement is to estimate the earnings that
individuals can expect over their work lives. This relation is referred to in the labour literature
as an age-earnings profile. A profile depicts the pattern of earnings of a cross-section of
workers at each age level. We then look for significant differentials in age-earnings profiles
between comparable workers in manufacturing and services. We interpret the results of this
approach to represent the earnings potential of typical service and manufacturing workers
over their work lives. This interpretation rests on the assumption that the behaviour of
individuals and labour market conditions affecting their wages do not vary significantly across
cohorts. Although this assumption may be open to question, the approach provides a starting
point for analysis (13).
Two approaches are pursued here. The first is to chart the age-earnings profiles of men and
women in the service and manufacturing sectors using hourly earnings. The second is to
estimate a wage equation that controls for worker characteristics and then plot the predicted
hourly earnings at each age level. The second approach constructs a statistically typical
person and thus plots the hourly wages of workers who are observationally identical except
for their age.
As shown in Figure 7 (located in the appendix at the end of the chapter), hourly earnings
increase throughout most of a worker’s career, regardless of whether the worker is male or
female, or in the service or manufacturing industries. The most widely cited reason for an
upward-sloping age-earnings profile is the accumulation of human capital through on-the-job
training. Other explanations include workers’ commitment to a firm by accepting low pay
early in their career in exchange for high pay later on. The age-earnings profile is steeper for
men than for women. Men and women start out with roughly the same pay, between the ages
of 18 and 22. However, men’s wages increase more rapidly than women’s wages, presumably
because women are less attached to the workforce due to their choice to assume household and
(13) Our analysis is an extension of the approach developed by Eberts and Groshen (1990) to compare the wage
growth opportunities in the service sector versus the manufacturing sector.
163
The job creation potential of the service sector in Europe
child-rearing responsibilities and thus do not accumulate as much human capital as men at any
given age.
For this study, the noteworthy pattern is that age-earnings profiles are roughly comparable for
men in manufacturing and services and for women in the two industries. Wages for young
entry-level workers are generally the same regardless of the gender of the worker or the
industry they work in. Earnings in the service and manufacturing sectors follow the same path
over the life of the worker. That is, wages increase for both service and manufacturing
workers at roughly the same rate. The wage differential between manufacturing and services
persists throughout a person’s work life. However, which industry claims a wage premium
differs by gender. For men, manufacturing workers are paid more than service workers. For
women, service workers are paid more than manufacturing workers. This relation holds true
for both years, although the difference is more pronounced in 1984.
The stylised age-earnings profiles, produced by estimating a wage for a statistical worker,
accentuates the differences and similarities between these curves in services and
manufacturing. As shown in Figure 8 in the appendix, entry-level wages for young workers
in 1984 are the same and the wages within manufacturing and services follow the same path.
However, the paths are different for men and women. As with the unadjusted data, service
wages are below manufacturing wages for a man’s entire work life. For women, the opposite
is true — service wages are higher than manufacturing wages. Wages in 1997 yield a similar
pattern, with one notable exception. The wages of men in the services start out higher than
wages of their counterparts in manufacturing. However, the curve for service wages is
shallower than that of manufacturing wages, and thus manufacturing wages eventually catch
up with service wages by the age of 36. After that point, wages in services and manufacturing
follow roughly the same path.
The results of the age-earnings profile indicate that service workers can expect wage growth
over their work life similar to manufacturing workers. These findings suggest that the service
sector does not create dead-end jobs. Furthermore, women can find higher wages in services
than in manufacturing, while men find a wage premium in manufacturing jobs.
1.8. Growth potential of a service sector economy
From a regional perspective (whether it is Member States of the EU or States within the United
States), the question is what is the potential for growth within the service sector to spawn
overall growth. This question is particularly pertinent for those areas that have lost a substantial
number of manufacturing jobs. Manufacturing is considered the quintessential export
industry—selling goods outside the area that in turn brings income into the local economy.
Exports thus create a multiplier effect, which increases the economic growth of the area.
Groshen (1987) addresses the issue of whether services have become a source of export-led
growth. Examining the Great Lakes region, she finds that services are exported outside a
region, and can be considered a viable component of the economic base. As of 1982, the time
164
The nature and determinants of service sector growth in the United States
period used in her analysis, there was apparently less trade in producer services than in
manufacturing. Furthermore, the exporting activities of services comprised a smaller
percentage of total employment within all United States MSAs than the export activities of
manufacturing. Based on calculations of export activity within MSAs, Groshen finds that
service sectors with high and moderately high export activity account for 1.5 % of total
employment in these MSAs. Manufacturing industries with high and moderately high export
activity, on the other hand, account for 5.5 % of total employment. The finding that
manufacturing has 3.7 times more export activity than services may be an upper boundary.
International trade figures for the United States show that the ratio of goods exports to services
exports was about 3.3 in 1982. However, one would expect the ratio to be larger for trade
within the United States since the effective distance that service exports may be ‘shipped’ may
be shorter than the effective distance for the shipment of goods exports. Obviously, more
work on the export nature of services needs to be pursued.
If services are exportable outside the region, then what effect does such activity have on the
local economy? Groshen (1993) also considers the economic impact of an increase in
expenditures or employment in services upon a regional economy. Considering regional
multipliers for Ohio, she finds that bringing to the region a business service provider that
generates USD 1 million in annual revenue creates 41 additional jobs. In contrast, bringing in
a tyre manufacturer that does the same amount of business creates only 23 more jobs. She also
finds that the service provider augments earning power to a greater extent than does the
manufacturer. For each additional dollar earned by the service firm, household earnings rise
by 89 cents, versus 54 cents for the manufacturer. She reports that a comparison of aggregated
services and manufacturing exhibits the same pattern as shown in the simple example of single
industries.
Part of the difference in economic impacts is due to the more labour-intensive nature of
services. Countering this effect is the lower indirect effect generated by services than by
manufacturing. One additional employee in services induces other firms to hire 0.8 additional
positions. In manufacturing, one additional employee results in other firms hiring 1.5 new
workers. Nonetheless, the net effect of raising service production by an additional dollar
yields more new positions in other industries than does the same rise in manufacturing sales.
Analysing the comparative economic impact of a USD 10 million increase in sales in business
services and motor vehicle production in the State of Michigan yields results similar to those
obtained by Groshen. As shown in Table 21, the increase in business service sales generates
2.3 times as many jobs as created by an identical increase in motor vehicle sales, even through
the State’s motor vehicle industry has a much larger employment multiplier. Each new job on
the auto assembly line generates an additional four jobs in the State (14). On the other hand,
each position created in the business service sector supports only another 0.6 jobs in the State.
(14) It should be noted that this is a very high employment multiplier. Most employment multipliers, even in
manufacturing industry, stay below 3. However, the State of Michigan houses a very large number of auto
suppliers.
165
The job creation potential of the service sector in Europe
Table 21. Economic impact of USD 10 million in new sales in the State of Michigan
Motor vehicle production
Business services
Total employment
— Direct
— Indirect
— Employment multiplier
Earnings per worker
Personal income
145 jobs
29 jobs
116 jobs
5.0
USD 52 540
USD 6.68 million
337 jobs
207 jobs
130 jobs
1.6
USD 36 824
USD 10.8 million
Gross State product
USD 8.96 million
USD 12.0 million
Source: W.E. Upjohn Institute, REMI model.
Since lower productivity in business services is lower than that in manufacturing —
USD 48 310 sales per worker in business services compared to USD 334 830 in sales per
worker in the motor vehicle industry — the overall earnings per worker impact of a jump in
business services is nearly 30 % lower than for the same increase in motor vehicle sales.
Nevertheless, the sector’s impact on the State’s personal income and gross State product is
higher by 61.7 % and 33.9 % respectively.
From an economic development perspective, the decision regarding which type of firm to
pursue depends upon the cost of nurturing a service firm versus a manufacturing firm, each
generating the same revenue. It also depends upon the effect of the local economic market
conditions on business creation. Our results in the previous section suggest that the creation
of service jobs is affected by factors such as wage rates, population size, and overall economic
activity, which are the same factors that affect manufacturing. Therefore, the preliminary
inquiry into this issue of the economic development potential of services suggests that service
sector businesses can stimulate a local economy and the magnitude of this effect is likely to
be as large if not larger than that generated by manufacturing firms. Moreover, previous
results from the analysis of MSAs suggest a tight relation between services and
manufacturing. In this light, services and manufacturing should be considered complements,
not substitutes, at least from a regional perspective.
1.9. Conclusion
The chapters in this volume explore the job creation potential of the service sector in Europe.
A motivating factor for this discussion is the slow pace of total employment growth in the EU
during the last decade or so, particularly when compared to the robust employment growth in
the United States. Both the EU and the United States have experienced significant growth in
service producing jobs during this time, but the share of service employment in the EU has
lagged behind that of the United States. The gap between the two economies in the size of the
service sector has raised questions about the future growth prospects for service sector
employment and ultimately total employment in the EU. Even more fundamental is the
concern about the growth potential of an economy that increasingly relies upon services.
166
The nature and determinants of service sector growth in the United States
The purpose of this chapter is to provide an overview of the service sector in the United States.
We address several basic questions: What are the determinants of service sector employment,
what is the relationship between services and manufacturing, what is the wage structure of
service jobs, and what is the potential for the service sector to stimulate overall employment
growth? Answers to these questions go beyond the scope of this chapter, but our analysis
provides some insight into these issues.
Much of our analysis focuses on business services, as they have achieved remarkable
employment growth during the past several decades. Business services are also considered
because they are closely related to the restructuring in the manufacturing sector and provide
intermediate inputs to manufacturers. Consequently, factors contributing to this tremendous
employment gain include the outsourcing of non-core activities by manufacturers. Not only
are the United States manufacturers losing share in many service-related occupations to the
service sector; the nation’s business and professional sectors are gaining share from
manufacturers in several production-related occupations, such as, machine operators and
assemblers. These findings suggest that temporary and lease employment agencies are
increasingly being called upon to meet the needs of manufacturers. Still, while evidence
indicates that structural change in manufacturing is a contributing factor to the growth in
service employment, its impact is small relative to the growth in the demand for services by
other services.
The lack of productivity growth in services accounts for a significant portion of the gap in
employment growth between services and manufacturing. In fact, as growth in service output
and manufacturing output are roughly the same, the difference in employment growth
between the two sectors is almost entirely the difference in productivity growth. This raises
the issue of the ability of services to support innovative activities that are needed to sustain
future productivity growth. It may be the case that while service sector productivity lags
behind manufacturing productivity, innovations in manufacturing are increasingly dependent
upon service activities such as engineering and research and development. When these
functions were performed internally by firms, the lag in productivity went unnoticed.
However, as these services are outsourced and sold in the market, the tendency to measure
productivity gains in these sectors, however crude and inaccurate, has raised concerns about
lagging service sector productivity growth. However, researchers have suggested that service
sector productivity should be measured by how effective the services are in increasing the
productivity of those activities that use these services as inputs. This approach relates closely
to the standards typically used when these services are provided internally.
Observing variations in service and manufacturing employment across metropolitan areas
offers some insights into the factors that contribute to their growth. Service employment is
concentrated in large metropolitan areas. The 78 largest metropolitan areas house 50 % of the
United States total service employment, while accounting for only 43 % of the nation’s total
employment. These metropolitan areas accounted for the same proportion of total
employment and service growth from 1985 to 1997. Based on observations of the 78 largest
metropolitan areas, our research suggests that employment in the services expand in regions
167
The job creation potential of the service sector in Europe
that have a large and dynamic manufacturing base, a low rate of unionisation, low
manufacturing wage rates and a relatively educated labour force.
Although the lack of productivity growth is a contributing factor to the employment growth
in services, wages and wage growth in services are becoming more comparable to those in
manufacturing. The differential for the median wage in services has declined from 87.6 % of
goods-producing sector wages to 93.1 % from the mid-1980s to the mid-1990s. Lesseducated workers still find better paying jobs in manufacturing; however, on average, service
workers can expect the same wage growth over their work life as manufacturing workers.
Finally, the service sector is becoming an increasingly important component of the economic
base of many regions in the United States. Services are no longer limited to a ‘non-base’
function that simply recirculates revenues brought into the region through the sale of
manufacturing goods. The regional exportation of services, such as health, financial and
professional services, can have a similar impact on the region’s economic base, as filling a
new order in manufacturing. Equally important, our research shows that manufacturing and
services activities are complementary. Hence, regional economic development efforts may
become more robust by creating a healthy economic environment for both its manufacturing
and service sectors.
Based upon these observations of the United States service sector, several policy
recommendations follow. First, as services become more exportable, it is important to extend
trade liberalisation measures to service firms that are given to manufacturers. When services,
such as banking and telecommunications, were more localised, there was little need to protect
local businesses from outside competition. However, as these businesses come under more
intense competition, the temptation to protect these entrepreneurs from the rigors of the
market should be carefully weighed against the benefits of more liberalised trade.
Second, the close relationship between restructuring within manufacturing and growth in the
service sector emphasises the need for flexibility in the labour market. Restrictions, either
explicit or implicit, on the mobility of workers may impede the ability of firms to adjust to
changing market demands and retard the growth potential of the service sector. Liberalising
labour adjustment within firms may improve the efficiency of the manufacturing sector by
promoting more efficient use of resources. Accommodating the move toward outsourcing
services may also increase the scope and efficiency of the service sector.
Third, the ability of service firms to generate as much additional employment as
manufacturing firms changes their role in promoting economic development. Instead of
focusing primarily on manufacturing, economic development efforts can also include service
firms. The findings that growth in service employment is sensitive to some of the same
location factors as manufacturing also emphasises that service firms can choose their location
and are not limited to a local market. Therefore, metropolitan areas must remain competitive
with respect to wage rates and the cost of other inputs used by business services.
168
The nature and determinants of service sector growth in the United States
Fourth, the greater role of business expansions for services than for manufacturing calls for a
shift in the approach taken to attract and retain businesses in a region. Instead of concentrating
on attracting new firms, perhaps more attention should be paid to retaining existing firms,
since employment growth appears to come more from expansions than openings.
Furthermore, as the turnover rate is higher for services than for manufacturing, it is important
to find ways to keep appraised of the problems facing service firms in case there are measures
that can be taken to keep the firm from leaving or going out of business.
Appendix
Table 1. List of metropolitan areas used in the analysis
Metropolitan areas
Population 1990
Akron, OH
Albany-Schenectady-Troy, NY
Albuquerque, NM
Allentown-Bethlehem-Easton, PA
Atlanta, GA
Augusta-Aiken, GA-SC
Austin-San Marcos, TX
Bakersfield, CA
Bergen-Passaic, NJ
Birmingham, AL
Charleston-North Charleston, SC
Charlotte-Gastonia-Rock Hill, NC-SC
Chicago, IL
Cleveland-Lorain-Elyria, OH
Columbia, SC
Columbus, OH
Corpus Christi, TX
Dallas, TX
Dayton-Springfield, OH
Des Moines, IA
Detroit, MI
El Paso, TX
Fort Lauderdale, FL
Fort Wayne, IN
Fort Worth-Arlington, TX
Fresno, CA
Grand Rapids-Muskegon-Holland, MI
Greensboro-Winston-Salem-High Point, NC
Hamilton-Middletown, OH
Houston, TX
Huntington-Ashland, WV-KY-OH
Jackson, MS
Knoxville, TN
Lakeland-Winter Haven, FL
Lancaster, PA
Las Vegas, NV-AZ
Lexington, KY
Little Rock-North Little Rock, AR
Los Angeles-Long Beach, CA
657 575
874 304
480 577
686 688
2 833 511
396 809
781 572
543 477
1 278 440
907 810
506 875
1 162 093
6 069 974
1 831 122
453 331
1 377 419
349 894
2 553 362
951 270
398 928
4 382 299
591 610
1 255 488
363 811
1 332 053
667 490
688 399
942 091
291 479
3 301 937
312 523
395 396
604 816
405 382
422 822
741 459
348 428
513 117
8 863 164
Metropolitan areas
Population 1990
McAllen-Edinburg-Mission, TX
Memphis, TN-AR-MS
Miami, FL
Middlesex-Somerset-Hunterdon, NJ
Minneapolis-St. Paul, MN-WI
Modesto, CA
Montgomery, AL
New Orleans, LA
New York, NY
Newark, NJ
Norfolk-Virginia Beach-Newport News, VA-NC
Oklahoma City, OK
Omaha, NE-IA
Pensacola, FL
Philadelphia, PA-NJ
Phoenix-Mesa, AZ
Pittsburgh, PA
Portland-Vancouver, OR-WA
Raleigh-Durham-Chapel Hill, NC
Riverside-San Bernardino, CA
Rochester, NY
Sacramento, CA
Saginaw-Bay City-Midland, MI
St Louis, MO-IL
San Antonio, TX
San Diego, CA
Santa Barbara-Santa Maria-Lompoc, CA
Seattle-Bellevue-Everett, WA
Shreveport-Bossier City, LA
Stockton-Lodi, CA
Syracuse, NY
Tacoma, WA
Tampa-St. Petersburg-Clearwater, FL
Toledo, OH
Tucson, AZ
Tulsa, OK
Vallejo-Fairfield-Napa, CA
Wichita, KS
Wilmington-Newark, DE-MD
383 545
981 747
1 937 094
1 019 835
2 464 124
370 522
292 517
1 238 816
8 546 846
1 824 321
1 396 107
958 839
618 262
344 406
4 856 881
2 122 101
2 056 705
1 239 842
735 480
2 588 793
1 002 410
1 481 102
399 320
2 444 099
1 302 099
2 498 016
369 608
1 972 961
334 341
480 628
659 864
586 203
2 067 959
614 128
669 880
708 954
451 186
485 270
578 587
169
The job creation potential of the service sector in Europe
Table 4. Employment change, 1984–97
Industries
Agricultural services
Mining
Construction
Durable goods
Non-durable goods
Trans. commun.
and public utilities
Wholesale trade
Retail trade
FIRE
Business and repair services
Personal services
Entertainment
and recreation services
Professional
and related services
Public administration
170
1984
Employment change due to:
1997
national
change in change in employment
growth
occupational occupational change
mix
share
2 267 952 2 241 043
1 035 250
686 753
6 424 606 7 282 634
13 233 498 12 704 758
9 350 819 8 572 726
432 870 – 643 727
183 948 – 26 910
197 592
– 26 338 – 519 750 – 348 496
1 226 225 – 519 098
150 901
858 028
2 525 796 – 992 853 – 2 061 682 – 525 740
1 784 733 – 1 202 728 – 1 360 098 – 778 094
7 560 464 9 257 971
4 159 462 4 555 180
19 029 148 23 330 266
6 626 407 8 140 467
5 046 556 7 691 703
4 308 415 4 043 912
1 443 019
793 891
3 631 975
1 264 741
963 205
822 320
189 054
74 364
583 851
308 523
– 99 893
– 700 081
2 798 699
284 421
153 083
871 018 1 308 522
22 458 006 31 158 266
5 578 760 6 827 100
4 286 420
1 064 783
1 892 449
983 408
2 521 392 8 700 261
– 799 852 1 248 340
1 490 177
65 435
– 472 537
85 293
– 59 204
1 781 836
– 386 742
1 697 507
395 718
4 301 118
1 514 060
2 645 147
– 264 502
The nature and determinants of service sector growth in the United States
Table 5. Outsourced occupations in manufacturing — employment loss in manufacturing due to its
losing occupational share
Occupations
Clerical and administrative support
Computer system analysis
Transportation occupations
Managers and administrators (salaried)
Production inspectors and testers
Freight stock, material handlers
Other professional specialities
Technicians, other
Machine operators, tenders
Material moving
Secretaries, stenographers and typists
Management-related occupations
Accountants and auditors
Cleaning and building
Construction trades
Engineers
Fabricators and assemblers
Natural scientists and mathematicians
Engineering and science technicians
Other precision workers
Employment loss
Percentage of 1994
employment
473 346
– 298 755
– 214 396
– 212 368
– 185 058
– 179 497
– 166 381
– 148 487
– 130 939
– 127 153
– 120 115
– 115 076
– 110 874
– 105 981
– 97 207
– 94 685
– 88 901
– 86 176
– 84 916
– 33.1
– 250.9
– 44.4
– 15.2
– 22.5
– 50.2
– 37.0
– 62.9
– 2.5
– 25.6
– 16.5
– 34.6
– 43.6
– 30.5
– 35.7
– 10.8
– 5.5
– 44.7
– 15.2
– 83 685
– 8.7
171
The job creation potential of the service sector in Europe
Table 6. Employment gains in business services by occupation due to increases in occupational share
Occupations
Clerical and support occupations
Computer system analysis
Managers and administrators (salaried)
Sales reps, commodities and financial
Cleaning and building services
Other specified handlers and cleaners
Labourers, excluding constuction
Mechanics and repairers
Engineers
Machine operators, tenders
Fabricators and assemblers
Technicians, other
Protective services
Production inspectors and testers
Other professional specialities
Computer equipment operators
Engineering and science technicians
Teachers, excluding post-secondary
Financial record processing occupations
Construction trades and extractive occupations
Accountants and auditors
172
Employment
Percentage of 1994
employment
326 386
309 480
194 103
177 382
170 015
119 742
83 418
74 167
68 417
67 895
63 610
63 356
59 520
56 736
51 555
31 967
26 013
23 875
21 761
20 334
62.2
883.7
42.4
112.4
35.2
96.0
129.9
9.9
159.5
44.6
56.2
77.3
19.1
319.7
41.7
39.3
33.3
268.5
16.2
61.2
16 563
35.3
The nature and determinants of service sector growth in the United States
Table 7. Employment gains in professional services by occupation due to increases in occupational share
Occupations
Managers and administrators (salaried)
Secretaries, stenographers and typists
Clerical and support occupations
Other professional specialities
Financial-record-processing occupations
Management-related occupations
Health service occupations
Technicians, other
Accountants and auditors
Engineers
Computer system analysts
Personal service occupations
Machine operators, tenders
Natural scientists and mathematicians
Mechanics and repairers
Engineering and science technicians
Construction trades and extractive occupations
Health assessment and treating occupations
Sales reps, commodities and financiers
Fabricators, assemblers and hand workers
Protective services
Employment
Percentage of 1994
employment
424 776
384 222
358 694
265 282
252 649
242 548
138 148
105 256
103 330
84 777
81 973
81 147
74 364
72 773
68 344
57 558
54 166
52 406
44 282
38 075
28.9
20.4
16.9
22.7
64.2
140.8
7.5
38.2
39.5
36.8
361.5
14.6
55.6
52.6
61.2
23.6
61.8
3.0
1 285.2
207.2
36 011
34.0
173
174
0.726 **
0.809*
0.05
0.263
0.07
0.958
Business
services
– 0.262 *
0.546 **
Services
A
0.12
0.688 **
0.023 **
– 0.018
– 0.304 **
0.224
Services
B
0.10
1.55 **
0.057 ***
0.010
0.695 **
0.343
Business
services
0.31
0.631 ***
– 0.003
– 0.029 *
– 0.008 ***
– 0.031 *
– 0.309 **
0.139
Services
C
0.23
1.56 ***
0.024
– 0.035
– 0.006
– 0.125 ***
0.630 *
0.257
Business
services
0.37
0.531 **
– 0.007
– 0.018
– 0.007 **
– 0.045 **
– 0 .0001 **
4.1e-09 **
– 0.0006
0.0017
– 0.281 **
0.385
Services
0.556 *
0.067
Business
services
0.41
0.44
Mean
1.75 ***
0.21
0.013
10.8
– 0.104 **
– 9.8
0.0002
15.1
– 0.162 ***
6.22
0.0002
143348
– 6.82e-09 *
0.004
59.3
0.020 **
23.3
0.337
0.445
0.30
D
NB: * (** ***) indicates statistical significance at the 0.10 (0.05, 0.01) level.
Dependent variables are the percentage change in the share of service employment between 1985 and 1997 for services and business services.
Source: Selected census data.
Openings share of job creation:
services
Openings share of job creation:
manufacturing
Manufacturing’s share
of employment 1985
% of job creation 1989-92
% of job destruction 1989-92
% of employees unionised
Manufacturing hourly wage
Per capita income
Per capita income squared
% of high-school graduates
% of college graduates
Dependent variable
Business services
Adjusted R-square
Variable
Model
Table 13. Percentage change in service sector share of total employment, selected MSAs, 1985–97
The job creation potential of the service sector in Europe
The nature and determinants of service sector growth in the United States
Table 14. Percentage change in service employment due to openings and expansions,
selected MSAs, 1989–92
Model
A
B
Variable
Openings Expansions
MSA population, 1990
Population squared
Manufacturing job
creation (% change)
Manufacturing job
destruction (% change)
Share of services, 1985
% employment unionised
Metropolitan areas
unemployment rate
Manufacturing hourly wage
% completing high school
% completing college
Manufacturing establishment size
8.16e-07 ** 2.39e-08
– 8.11e-14 ** – 1.54e-14
0.350 ***
0.119
– 10.1 **
0.23
0.224 ***
0.061
– 2.33
0.13
C
Openings Expansions
9.19e-07 *** 9.8e-08
– 7.99e-14 ** – 1.5e-14
0.226 ***
0.109
– 5.53 *
– 0.116 ***
0.143 **
Openings Expansions
9.09e-07 *** – 6.24e-08
– 6.80e-14 **
6.59e-15
0.279 ***
0.169 ***
0.054
–0.104
0.145
– 7.45 ***
– 0.075 *** – 0.137 ***
0.035
– 1.99
– 0.081 ***
– 0.087
– 0.070
0.47
0.31
– 0.160 *
0.123
0.036
– 0.045
0.50
0.072
– 0.099
0.048 **
0.048 *
0.33
NB: * (** ***) indicates statistical significance at the 0.10 (0.05, 0.01) level.
Dependent variables are the percentage change in service employment due to openings and
expansions, 1989–92.
Source: Selected census data.
175
The job creation potential of the service sector in Europe
Table 15. Weekly wages in the eight major industries, 1984–86
Mean
Goods-producing
Construction
Manufacturing
Service-producing
Retail trade
Narrow services
Wholesale trade
FIRE
Public administration
Transportation and utilities
10th percentile
Median
90th percentile
Share of
employment
USD
USD
%
USD
%
USD
%
%
656
623
665
592
440
565
657
657
687
770
242
232
247
208
161
201
258
265
297
317
100.0
95.9
102.1
86.0
66.5
83.1
106.6
109.5
122.7
131.0
563
528
564
493
344
475
549
499
618
713
100.0
93.8
100.2
87.6
61.1
84.4
97.5
88.6
109.8
126.6
1 142
1 119
1 147
1 061
817
1 004
1 159
1 198
1 155
1 218
100.0
98.0
100.4
92.9
71.5
87.9
101.5
104.9
101.1
106.7
32.2
6.9
25.3
67.8
13.6
28.1
4.5
7.1
6.1
8.3
Table 16. Weekly wages for workers aged 25 to 30, 1984–86
Mean
Goods-producing
Construction
Manufacturing
Service-producing
Retail trade
Narrow services
Wholesale trade
FIRE
Public administration
Transportation and utilities
10th percentile
Median
90th percentile
Share of
employment
USD
USD
%
USD
%
USD
%
%
583
587
582
531
448
512
577
578
585
647
245
244
246
226
186
220
282
285
304
309
100.0
99.6
100.4
92.2
75.9
89.8
115.1
116.3
124.1
126.1
525
516
530
470
377
451
516
488
527
602
100.0
98.3
101.0
89.5
71.8
85.9
98.3
93.0
100.4
114.7
976
1 014
958
867
772
832
911
918
891
1 010
100.0
103.9
98.2
88.8
79.1
85.2
93.3
94.1
91.3
103.5
31.8
7.6
24.2
68.2
14.9
27.5
5.1
8.1
4.9
7.7
Table 17. Weekly wages for workers with only a high-school diploma, 1984–86
Mean
Goods-producing
Construction
Manufacturing
Service-producing
Retail trade
Narrow services
Wholesale trade
FIRE
Public administration
Transportation and utilities
176
10th percentile
Median
90th percentile
Share of
employment
USD
USD
%
USD
%
USD
%
%
615
620
613
500
414
428
577
508
577
721
252
244
253
197
161
179
257
247
275
316
100.0
96.8
100.4
78.2
63.9
71.0
102.0
98.0
109.1
125.4
542
535
545
422
338
373
507
422
527
702
100.0
98.7
100.6
77.9
62.4
68.8
93.5
77.9
97.2
129.5
1 035
1 099
1 014
891
741
722
958
861
962
1 126
100.0
106.2
98.0
86.1
71.6
69.8
92.6
83.2
92.9
108.8
34.5
7.7
26.8
65.5
16.3
21.9
4.7
7.0
5.9
9.5
The nature and determinants of service sector growth in the United States
Table 18. Weekly wages in the eight major industries, 1995–97
Mean
Goods-producing
Construction
Manufacturing
Service-producing
Retail trade
Narrow services
Wholesale trade
FIRE
Public administration
Transportation and utilities
10th percentile
Median
90th percentile
Share of
employment
USD
USD
%
USD
%
USD
%
%
685
633
702
657
450
657
726
827
735
775
236
236
236
210
165
205
246
273
327
300
100.0
100.0
100.0
89.0
69.9
86.9
104.2
115.7
138.6
127.1
537
527
538
500
346
498
538
547
630
673
100.0
98.1
100.2
93.1
64.4
92.7
100.2
101.9
117.3
125.3
1180
1 079
1 215
1 132
790
1 131
1 215
1 377
1 194
1 226
100.0
91.4
103.0
95.9
66.9
95.8
103.0
116.7
101.2
103.9
27.2
6.5
20.7
72.8
14.3
32.6
4.2
7.0
6.5
8.1
Table 19. Weekly wages for workers aged 25 to 30, 1995–97
Mean
10th percentile
Median
USD
USD
%
USD
%
Goods-producing
Construction
Manufacturing
Service-producing
Retail trade
Narrow services
Wholesale trade
FIRE
Public administration
551
550
552
532
421
544
574
614
583
231
240
224
212
179
211
250
273
295
100.0
103.9
97.0
91.8
77.5
91.3
108.2
118.2
127.7
466
472
462
453
360
462
492
483
519
Transportation and utilities
584
269
116.5
527
90th percentile
Share of
employment
USD
%
%
100.0
101.3
99.1
97.2
77.3
99.1
105.6
103.6
111.4
923
913
932
865
699
865
920
957
911
100.0
98.9
101.0
93.7
75.7
93.7
99.7
103.7
98.7
25.8
6.9
18.9
74.2
16.8
32.4
4.3
8.0
5.9
113.1
962
104.2
6.8
Table 20. Weekly wages for workers with only a high-school diploma, 1995–97
Mean
Goods-producing
Construction
Manufacturing
Service-producing
Retail trade
Narrow services
Wholesale trade
FIRE
Public administration
Transportation and utilities
10th percentile
Median
90th percentile
Share of
employment
USD
USD
%
USD
%
USD
%
%
583
613
572
501
411
426
624
581
576
707
238
255
236
197
172
179
247
256
277
288
100.0
107.1
99.2
82.8
72.3
75.2
103.8
107.6
116.4
121.0
500
527
490
405
332
363
492
449
511
649
100.0
105.4
98.0
81.0
66.4
72.6
98.4
89.8
102.2
129.8
984
1 013
962
833
708
692
925
865
911
1 082
100.0
102.9
97.8
84.7
72.0
70.3
94.0
87.9
92.6
110.0
33.8
8.7
25.1
66.2
17.4
23.7
4.6
6.1
5.1
9.2
177
The job creation potential of the service sector in Europe
Figure 7. Age earnings profile for men and women in manufacturing and services, 1984 and 1997
Figure 8. Age-earnings profile for men and women in manufacturing and services, 1997 and 1984
(mean hourly wage for each group)
178
The nature and determinants of service sector growth in the United States
Figure 9. Percentage change in service industry share, 1985–97, for metropolitan areas in the sample
Figure 10. Percentage change in service industry employment; 1985–97, for metropolitan areas in the
sample
179
Chapter 2
Determinants of business and personal
services: Evidence from the German regions
Dietmar Dathe and Gunther Schmid
2.1. Introduction
The objective of this study is to identify the determinants of employment in business services
and personal services. As research strategy we use the variations in the structure and dynamics
of regional employment within Germany in the period of 1977 to 1998. The observation of
great differences in employment or unemployment performance at this level is not a German
specificity; it can be found within other Member States of the European Union as well as in
the United States and Japan. The German case, however, is interesting for its massive service
gap compared to Anglo-Saxon (United States, United Kingdom) and Scandinavian countries
(Denmark, Sweden) in general, and in personal and business services in particular (Freeman
and Schettkat 1999).
In neo-classical economics, the regional dimension plays no genuine role in the search for
universal rules. In the new institutional economics, however, regions became at least
instrumentally interesting as cases of variation to test hypotheses related to institutions of
wage formation. A prominent case is the ‘wage curve’ which forecasts on the basis of
bargaining theory or efficiency wage theory a persistent negative relationship between
regional unemployment rates and wages. An elasticity of – 0.1 was found and confirmed for
several countries which means that a doubling of the regional unemployment rate leads to
10 % lower average wages (Blanchflower and Oswald, 1994). Although the ‘iron law’ of the
wage curve is still contested (Blien, 1996), the policy message is clear: If such regional
unemployment equilibria exist, they can only be tackled by changing the institutional setting
of regional actors. The likelihood of such disequilibria is confirmed by studies which show
that inter-regional employment or unemployment variations can largely be explained by
regional-specific (namely, cultural, institutional and political) factors (1).
(1) See, for example, de Kooning and Mosley, 2000; Clark 1998; Miegel, Grünewald and Grüske, 1991.
181
The job creation potential of the service sector in Europe
Comparative sociology, up to now, has relied almost completely on national regimes to
explain differences in the performance of service employment. An example is the widely
quoted typology of welfare state regimes by Esping-Andersen (1990) which relates job
performance to the degree of commodification. National regimes with a high degree of
commodification, which means the extent to which income security is directly linked to labour
market participation (private or public), will have a higher job performance than countries in
which the State provides social security to certain status groups independent of market
income. Since we find, nevertheless, enormous differences of job performance within the
same national regimes, implementation of national regulatory frameworks or additional
institutional factors must be taken into account to explain these variations.
There are, therefore, plausible reasons to start with two assumptions. First, the capacity of
wages to regulate labour markets is restricted, either due to behavioural limits (for example,
in information, spatial mobility or competition) or due to social norms or objectives (for
example, fair wages, equal opportunity, minimum income guarantee). Second, there is relative
autonomy of regional actors from framework conditions at the national level. This is not to
deny the determination power of national institutional frameworks, but to exploit a neglected
source of variation which gains increasing plausibility considering the globalisation process.
This process favours or fosters network relations which operate between markets (wage
regulation) and hierarchies (State regulation). They flourish usually in a specific regional
context which is not restricted to national boundaries (Castells,1996, Genosko, 1999).
The epistemological argument for these assumptions is based on the insight of evolutionary
theory that formalised institutional rules always lag behind informal institutional
arrangements (Boulding, 1978). The same institutional rules or games can be played
differently, depending on resources, capabilities, learning processes and cultures. Institutions
are always both: effective barriers but also challenges to be overcome or circumvented. In
addition, Germany is a Federal State in which the 16 Länder have not only their own cultural
traditions but also considerable political discretion to shape their economic and social
systems; education, especially, is subject to the Länderhoheit (sovereignty), and social
assistance (including often many long-term unemployed as clients) is under the genuine
responsibility of the municipalities. Finally, the assumption is widespread that agglomerated
areas, especially metropolitan cities, are heading the process of social and structural change;
thus, drawing attention to these particular types of regions may provide hints about the future
of work, especially service jobs (Sassen, 1991).
Most of our empirical analysis, therefore, will concentrate on a systematic comparison of
‘winners’ and ‘losers’ among the 11 largest agglomerated areas in west Germany. These
regions are technically based on the aggregation of public employment service districts and
constitute functionally integrated labour markets. They cover about 50 % of the total
employment population in west Germany. Due to their very specific problems related to the
reunification after 1989, east German regions and Berlin were not included in this study. The
non-agglomerated areas in west Germany serve as a case of reference for the agglomerated
areas.
182
Determinants of business and personal services: Evidence from the German regions
The structure of the chapter is organised as follows: We start from the obligatory ‘classics’ and
develop hypotheses based on a review of recent literature on business and personal services
(Section 2.2); the hypotheses are then tested with German data and compared with results of
other literature using regional variations (Section 2.3); and finally the results are discussed and
policy conclusions are drawn (Section 2.4).
2.2. Theoretical notes and observations on the service gap
The notion of post-industrial society (Bell, 1975) is linked to the ‘Three-Sector-Hypothesis’
(Fourastié 1969 (1954)). It predicts an evolutionary process from employment in agriculture
to manufacturing and finally to services. The driving forces in this theory are prices based on
productivity and demand based on hierarchical needs. It is important to note here, that
according to this theory — and against the mainstream of neo-classical economy — it is not
supply and demand that drive the evolutionary process but social and technical innovation.
Innovation induces growth in capital and labour productivity, a corresponding long-term
decline of prices and therefore an increase in real income. Information and knowledge bring
forward technical innovations first in agricultural production, where prices fall and cause a
saturation of needs and demand for ‘higher’ durable goods (Fourastié and Schneider, 1989).
Innovation shifts towards manufacturing, taylorist and fordist production increase
productivity in this sector, prices fall and demand as well as employment increases until a
relative degree of saturation is reached (so-called Verdoon’s law). Demand, then, shifts
towards ‘higher’ services. The increasing purchasing power through (further) falling prices of
primary and secondary goods can be used to buy services whose prices remain stagnant or
even rise as productivity gains in this sector remain restricted.
According to Fourastié, service employment can rise as long as the motor of technological
innovations drives further productivity gains in the primary and secondary sectors. Thus, the
basic message for politicians was clear from this story: induce and support technological
innovation as far as possible so that society can afford to pay for services. Because Fourastié
believed in technical progress, he optimistically predicted that at the end of the 20th century,
about 80 % of the labour force would be employed in the tertiary sector, and only 10 % in the
primary (agriculture) and the secondary (industry) sector respectively. Compared with other
speculations that often turned out to be crazy a few years later, this prediction from almost 50
years ago fits the reality quite well and seems, therefore, well founded. However, there are
three stylised facts that require reflection and fine tuning or revision of the theory. First,
agricultural employment declined even faster and more drastically than predicted; second, the
reverse is true for manufacturing employment which is still at double the level predicted (in
some countries, notably in Germany, even more); and, third, service employment in some
countries and in many regions could not compensate for the loss of jobs in the primary and
secondary sectors so that mass unemployment is persistent there.
Before we turn to the special case of Germany, we will briefly mention three strands of an
argument which temper Fourastié’s optimism without going into details or into a critical
discussion of these approaches. First, services can also, to some extent, be transformed
183
The job creation potential of the service sector in Europe
through innovation into (secondary) manufactured goods plus self-servicing if one thinks of
typewriters (now PCs), cars or washing machines (Gershuny, 1983). Second, since selfservicing and consumption of services require time, and since individual time budgets are
chronically restricted, expectations of job miracles in services should be rejected (Scharpf
1990). Third, there is the well-known theory of cost disease which predicts that, due to the
impossibility of productivity gains and due to the link of service wages to secondary wages,
prices of many highly valued services, especially cultural and personal services, will rise to
such an extent that many people will no longer be able to afford them (Baumol, 1967) (2).
Why, now, is there a service gap in Germany? Which additional theories can help us
understand the deviations from the core prediction of Fourastié? The following deliberation
regarding these questions will concentrate on the recent German debate.
2.2.1. The service gap I: The case of personal services in Germany
In international comparison, three stylised employment regimes can be distinguished (Schmid
1992, 1993): (1) Regimes in which wages are paid in accordance with productivity or market
value; that is, high-productivity jobs pay commensurately higher wages, while lowproductivity jobs (especially in the service sector) pay commensurately lower wages; targeted
welfare payments or kinds of negative income taxes provide social protection at the last resort.
This is the Anglo-Saxon model in which private (tertiary) service jobs compensate for the loss
of manufacturing (secondary) jobs. (2) Regimes in which every job should guarantee a
socially adequate income; in the private sphere, this guarantee is implemented by minimum
wages or family components of wages; jobs in the social sphere are created by the State and
financed from tax revenue or other contributions. This is the Scandinavian model in which
public service jobs take over the compensatory function for the historic decline of the
secondary sector. (3) Regimes in which new jobs with low productivity (especially in the
service sector) are not subsidised due to a high-skill, high-productivity and corresponding
high-wages strategy on the one hand, and non-interventionist employment policy on the other
hand. Low skill, less productive tasks are left largely to the informal sector or private
households to supply society with these services. Germany comes close to this model. It
‘combines (…) the weaknesses of the American and the Swedish models. We have almost as
few workers in private services as Sweden and exactly as few workers in the public sector as
the United States’ (Zukunftskommission 1998, p. 242).
2.2.1.1. Doubts about the empirical evidence of the service gap
Empirical estimates concerning the size of the service gap vary, in some cases considerably,
depending on the approach chosen (whether sectoral or functional) and the classification and
measuring method applied. Some authors use potential values derived from ‘United States
structures’ to determine the size of the service gap. Klös (1997) arrived at an employment
(2) See Tronti et al., 2000 for an extensive discussion and critic of this theory.
184
Determinants of business and personal services: Evidence from the German regions
potential totalling 6.8 million jobs on the basis of job density in the tertiary sector (persons in
employment per 1 000 inhabitants). Following ‘United States structures’, 3.8 million of these
jobs (distributive and household related services) would be suitable for the so-called lowskilled. The employment gap resulting from the minimum wage is somewhere in the region of
4.7 to 9.5 million, according to Klös, jobs that by definition would largely be on offer to lowskilled workers. Sesselmeier et al. (1996), on the other hand (but likewise basing their
calculations on the employment structure in the United States), identified an employment
potential of 2.8 million in Germany’s ‘low-wage sector’.
Haisken-DeNew et al. (1996), by contrast, challenge the existence of a quantitative service
gap on the basis of the share of services in total employment. Cornetz and Schäfer (1998),
using the same micro-data sources as Haisken-DeNew et al. (The United States: Current
Population Survey; Germany: socioeconomic panel), but a modified approach, arrived at a
different result. The Otto-Blume-Institute in Cologne and the German Institute for Economy
in Berlin (Schupp. et al., 1997) point to the under-reporting of marginal jobs below the
DEM 630 threshold (‘geringfügige Beschäftigung’ not liable to social security contributions).
They argue that a proper account of these primarily low-wage jobs would largely close the gap
between the United States and Germany (3). On the other hand it is clear that such a threshold
creates disincentives for employers to hire people for jobs in the range between this threshold
and the full social security threshold which is at 15 hours regular weekly working time with
an average monthly gross wage of about DEM 1 200. Regular part-time jobs between the
DEM 630 threshold and 15 hours (practically speaking between 10 and 15 hours), are heavily
under-represented (4).
2.2.1.2. Underdeveloped household related services
The German discussion about the insufficiently tapped employment potential of personal
services is focusing on ‘potentials for new services geared towards demand from private
households’, (Weinkopf, 1999, p. 195). According to the literature, this potential consists in
particular of household related services (Albach, 1989) or domestic services (Europäische
Kommission, 1995): These are understood to include both material services or housekeeping
— such as shopping, preparation of meals, care of clothing and living quarters — and personal
services, in so far as the latter are related to the supervision and care of persons living in the
household or who are associated with the family (5). Given that these activities have been
(3) In fact, the recent official adjustment of employment figures by the Federal Agency of Statistics (Statistisches
Bundesamt) added about 2 million more employees (from a reported 34 million to 36 million, that is, an
increase of about 6 %), mostly related to a more precise account of marginal jobs.
(4) See the box regarding the so-called DEM 630 jobs in Germany in Appendix 1.
(5) The classification ‘household related services’ is not used consistently in the literature. In order to distinguish
such services from enterprise-related services, all services availed of by private individuals are categorised
either as household related (Albach, 1989) or as personal services (Schettkat, 1996). Häußermann and Siebel
(1995) differentiate between household related services, personal services and consumer-oriented services. The
European Commission (Europäische Kommission 1995) subsumes domestic services (in addition to childcare,
new information and communications services, assistance for young people in difficulties and integration
services) under the generic term ‘services for daily life’. Given that it is impossible to clearly differentiate
household related services on the basis of the official statistics, the areas of trade, hotels and restaurants, the
public service, and social and personal services, for example (according to the OECD Labour Force Statistics),
are classified as services that benefit domestic turnover (cf. Zukunftskommission, 1998).
185
The job creation potential of the service sector in Europe
carried out to date largely in the form of (unpaid) housework (Meyer, 1997), the hope that new
employment potential can be tapped is based on the assumption of increasing efficiency
through the commercialisation and professionalisation of this informal housework
(Kommission, 1997). The European Commission (Europäische Kommission, 1995) traces the
(assumed) willingness to replace housework by external market-conforming services back to
structural changes in society; to the ageing of the population, the increased labour force
participation of women, the reductions in working time, the higher levels of education, and to
urbanisation. These ‘elementary activities’ (SOFI, 1999) or ‘elementary services’ (Kommission,
1997) are believed to constitute the potential jobs for the low-skilled (Klös, 1997;
Zukunftskommission 1998) (6).
Another aspect of the current debate concerns the term ‘services or service areas that are
associated with the household’, (Deitelhoff, 1999; Evers, 1999). This aspect is linked to
concepts favouring a return of employment policy to the communal level (SOFI, 1999) or the
exploitation of the local dimensions to development and employment (Europäische
Kommission, 1995). The types of enterprise and employment discussed in this context range
from ‘household related small businesses’ and ‘private households as service enterprises’
(Albach, 1989), through local employment initiatives — understood as sponsors of ‘strategies
for new household related services and provisions that have a community orientation, are
business entities and seek to satisfy individual needs’ (Evers, 1999, p. 108) — to an
employment sector called ‘services for private households’ which is liable for social insurance
contributions (Weinkopf, 1996, 1997).
Gries and Birk (1999), on the other hand, trace the service gap back to the employment deficit
in Germany in the financial sector compared to the United States and find no particular
development shortfalls in the ‘low-wage sector’ or in personal services. The ifo-institute in
Munich also comes to the conclusion that the research and development activities in servicespecific information technologies (for management, administration, design, financing and
marketing) are very weak in Germany (Faust et al., 1999). They stress, however, the point that
the ‘tertiarisation’ of German manufacturing industries is far more advanced than official
statistics seem to show. These observations are worth examining more closely in theoretical
terms.
2.2.2. The service gap II: The case of business services in Germany
Business services (7), in very general terms, are services that enterprises carry out (primarily)
for other enterprises. The current debate is focused, not only in Germany, on so-called
(6) Because the current employment policy proposals concerning exploitation of the employment potential of
services are related almost exclusively to this area, these concepts in particular will be presented in the
following.
(7) Other terms in the literature are ‘producer services’ or ‘production oriented services’, and in the German
literature unternehmensnahe Dienstleistungen or produktionsorientierte Dienstleistungen. In a more narrow
sense, production-oriented services relate only to services for business in manufacturing (tangible) products. In
the following, however, we use the more general term including services to businesses in the service sector, for
example, advertisement for banks or private insurance companies.
186
Determinants of business and personal services: Evidence from the German regions
knowledge-intensive services (software and data-processing, financial services, legal,
economic and technical consulting, advertising and marketing). The conceptual basis for the
term ‘knowledge-intensive’ is the assumption that the ‘codification of theoretical knowledge’
(Bell, 1999, p. XXXIX), is the essential distinguishing feature of ‘post-industrial society’. The
starting point for inventions, innovations and their diffusion (with positive employment
effects) is not empirically inductive, but theoretically deductive knowledge. In other words, a
significant feature of the path leading to a (theory-based) knowledge society will be
interaction, or innovative milieus, between knowledge-intensive services and knowledgeintensive industrial sectors (8).
Estimates regarding the employment effects of an expansion in business or producer services
vary considerably. While researchers looking at the United States (Hummel, 1999) speak of a
hugely untapped employment potential, others (Baethge, 1999) point to the large potential for
rationalisation in information and communication technologies (in connection with their
continued expansion), and in modern services (by virtue of their increasing geographical
autonomy and tradability). Moreover, the employment trends within the business-service
sector are by no means found to be identical. Employment is rising in some branches while it
is falling in others (SOFI, 1999). Overall, the growth of business services has a much wider
margin of fluctuation than that of the manufacturing sector. Kaiser and Voß (1999) suspect, in
the light of the continuing significance of personnel costs, that service firms offering business
services react (and can react) more flexibly to demand fluctuations because the costs of high
personnel fluctuation are lower for them than for industrial enterprises, where there is a higher
degree of regulation. According to Licht et al. (1997), most business service firms in Germany
nonetheless expect positive employment trends in the coming years, or at least they plan to
employ more personnel. However, the positive expectations for employment expressly
exclude workers without a vocational qualification. What are the driving forces which foster
business services?
2.2.2.1. Interaction between employment in industry and in business services
Proponents of the supply theory believe that the growth of the service sector is the result of
increasing competitive pressure forcing industrial enterprises to offer high-quality products,
into which increasingly diverse services then flow (development, design, marketing, customer
support, etc.). The traditional value added chain is seen to be turning into a value added
alliance; because few enterprises today are able to efficiently carry out all value added
activities, the importance of a value added alliance or cooperative value added systems
(networks) is increasing. The result is that fewer value added operations are being carried out
in enterprises, and processes are being outsourced because they can be accomplished by
specialised service firms with better know-how and at a lower cost. The decision as to which
services can be provided internally and which bought on the market (or rather, which services
(8) In operative terms, we will measure ‘knowledge-intensive industries’ later on by an above-average share of
employees with an academic qualification and or above average expenditure on product and process innovation
(see Section 2.3).
187
The job creation potential of the service sector in Europe
are outsourced) depends on the ensuing costs for the enterprise. The advantages of external
services are not only the expertise and the higher degree of flexibility offered by private
service firms, but also the fact that enterprises that purchase external services are not subject,
in this case, to regulations regarding dismissal, working time and co-determination. The
competitive pressure of external suppliers, in addition, might reduce prices without affecting
quality. All this taken together reduces hourly labour costs for external compared to internal
services. The proponents of this approach logically conclude that the relatively low diffusion
of services in the Federal Republic of Germany, compared to the United States, for example,
is due to the high degree of regulation (for instance under labour law), which also applies to
the service sector characterised by a majority of small and medium-sized enterprises. The
strong dynamic observed nonetheless in business services is due to the existence of a
functioning market in this sector (Albach, 1989b), which would thrive even more, according
to the supply theorists, if there were further deregulation.
There is disagreement in the discussion as to whether the process of outsourcing influences
employment trends and as to whether failure to outsource results in a competitive
disadvantage on international markets. One reason is that the shortcomings of current
statistical data on business services prevent quantitatively accurate conclusions. According to
Schimmelpfennig (1999), utilisation of internal services in the manufacturing sector actually
increased between 1984 and 1995, outsourcing notwithstanding. This corresponds with an
American study of the 1970s and 1980s showing that outsourcing or ‘the unbundling
explanation accounts for a very small portion of the recent employment growth of producer
services industries’, (Tschetter 1987:31). This study explained the bulk of the growth
dynamics in business services by new services related to changes in final demand and to
changes in production due to new information technologies. So, if outsourcing occurs, it is
related to new services rather than to internal restructuring that would largely turn out as a zero
sum game in terms of employment.
For Häußermann and Siebel (1995), sectoral tertiarisation (due to outsourcing) is to an extent
no more than a statistical artefact; thus, differences in the sectoral structure of employment
should not necessarily be interpreted as an indication that services are less important. By
contrast, Strambach (1997) believes that the lower degree of outsourcing is evidence of an
institutional weakness in the German economic model: spin-off processes require greater
individual economic efficiency in the now independent service firms, while at the same time
they increase efficiency in the enterprises that have divested (theory of interaction). The
interaction theory expresses the idea that rationalisation and innovation in industry have
spillover effects for production-related services (especially high-quality services), which in
turn improve the competitiveness of the manufacturing sector. According to the extended
interaction theory, spillover effects are also generated within the service sector itself. In other
words, these enterprises not only function as demanders of technical and product innovations,
but also generate such innovations themselves (Wyckoff, 1996). The institutional weakness in
the area of business services is also seen to manifest itself in a lack of competitiveness on
international markets. An increase in service exports would lead to domestic employment
gains in the same way as a substitution of imported services by services produced at home
(Hild et al., 1999). This assessment seems to be supported by the facts: while the United States
188
Determinants of business and personal services: Evidence from the German regions
is achieving an international trade surplus of over USD 50 billion in knowledge-intensive
services, in 1998 Germany had a trade deficit of DEM 62 billion in this sector (9).
We need to ask whether the trade deficit is an indication of a decline in Germany’s innovative
strength (Zukunftskommission, 1998) or whether it expresses a different profile of innovative
strengths: ‘The decisive difference between Germany, the United States and the United Kingdom
is thus not the less evident high-tech sector, but the smaller (in sectoral terms — author’s note)
service sector’ (BMBF, 1999: 42). Supply and demand-based theoretical approaches are not
perforce mutually exclusive, rather they demonstrate the multi-dimensionality of services. This
complexity is embraced by approaches that combine the supply and demand aspects, which
assume the existence of a positive feedback effect: processes of rationalisation and adjustment to
changing competitive conditions result in increased demand for services, which in turn guarantee
efficient industrial production. This also means that increased utilisation of services by industry
does not necessarily prevent growing interaction between industrial and service enterprises
(interaction theory), rather the two can be parallel processes (O’Farell, 1995). The interaction
theory is also supported by the innovation hypothesis, according to which industrial demand for
input is the driving force behind structural change (10). Related to the interaction theory is the term
‘hybrid products’which refers to manufactured products including client-oriented services before
or after the production of or during the use of the product. ‘Hybrid products’ take into account
changing or rising consumer demands. Clients are more apt to buy ‘problem solving’systems than
individual products (Bullinger, 1998, Ganz, 1999).
2.2.2.2. Business services and regions
The theory of interaction is supported by the many contributions to discussions about
‘industrial districts’ and ‘local economic networks’. (11) These approaches incorporate the
levels of political and economic actors, institutional structure and region-specific economic
cultures in order to interpret regional differences (12). The success of such ‘industrial districts’
is seen above all in the specific (regional) mix of industrial and service enterprises, which
exert reciprocal and positive feedback effects and grow stronger on the basis of a cumulative
process (13). In this view, business services are considered extremely significant as catalysts of
(9) The negative German trade balance is primarily due to the fact that Germany is still ‘world champion’ in tourist
travel (a service import), but the above assessment still applies when travel is excluded from the equation.
(10) On balance, industrial enterprises purchased around DEM 250 billion worth of input from business-service
firms in 1997. Measured as an overall share of gross value added, the significance of this input alliance between
industry and service firms has increased considerably in recent years. While the manufacturing sector’s share
of GDP (without input) was 34.9 % and the share including input was 35.6 % in 1978, the two values had
diverged sharply by 1993; 25.8 % and 31.7 % respectively. In other words, the degree of interpenetration
between industry and business services has increased. For the year 2000, the manufacturing sector’s share of
GDP (including input) is expected to be 32 %, against only 23 % without input (see Daimler Chrysler Services,
1999).
(11) Exemplified by essays in Pyke and Sengenberger (1992), Camagni (1991) and Häußermann (1992); for an
overview Genosko (1999).
(12) This is also the framework or point of reference for political intervention and regulation. A summary of the
scientific debate can be found in Heinze and Schmid (1994).
(13) However, the possibility of a ‘negative spiral’ also exists (Grabher, 1994: 65). This is the case, for example, in
the Ruhr Valley, which was long considered the classic example of an ‘industrial district’, but has now become
a crisis zone because of entrenched and one-sided economic structures and adherence to traditional practices
despite changed conditions on the world market. The endogenous potential in the area must now be exploited
to develop new production networks (Kilper et al,. 1994).
189
The job creation potential of the service sector in Europe
structural change (Kilper et al., 1994; Brusco, 1992; Läpple, 1994; Begg, 1993). ‘Innovative
milieus’ or ‘innovative networks’ represent a relatively recent approach (on Germany, see, for
example, Beise et al., 1998; Läpple, 1989; Ronneberger, 1992). What these approaches have
in common is the question as to which factors give rise to conditions that foster innovation,
and the regional approach (milieu) is combined with the actor approach (network) to provide
an answer. The network consists of regionally based public and private research and
development institutions, political actors and enterprises that maintain successful — both
formal and informal — network relationships at the regional level. Innovative networks or
milieus are also supported by the outsourcing approach; by reducing their value added
activities, industrial enterprises become more dependent on a network of types of cooperation
that have as their objective expansion and diversification of the production and service
programme (Sauer and Döhl, 1994). In a sense, these concepts tie in with the term ‘industrial
atmosphere’ coined by Marshall (cf. Ronneberger, 1992) in connection with his theory of
industrial districts, which today can be translated as ‘innovative atmosphere’.
As regards the spatial effects of employment development, two theoretical approaches are
relevant here (Lichtblau and Rhein 1993). It is assumed in the more recent location theory that
industrial conurbations have exploratory advantages over peripheral or rural regions and can
thus evolve more dynamically. This is explained by the (positive) external effects that are
generated when many enterprises belonging to the same sector or technology domain are in
geographical proximity, and by the development of specific human capital, a particular
infrastructure and informal networks. By contrast, the factor equalisation theory (the old location
theory), sees regions converging because, compared to the dynamic centres, costs (wages, rent,
etc.) are lower and because there are increasing bottlenecks (transport, lack of qualified
manpower, etc.) in the high-density areas, so that enterprises are increasingly locating in the less
dense regions. An intervening factor is the political constitution. In Germany, one can speak of
a polycentric distribution of innovation potential, which is a considerably different situation to
that in the United States, but also to that in France, Italy and the United Kingdom (Beise et al.,
1999). This polycentric distribution of the innovation potential is mainly a consequence of
Germany’s federal structure, which has the effect of a much more balanced distribution of
universities and publicly funded non-university institutions than in other industrialised countries
(Beise et al., 1998). However, despite this relatively balanced regional distribution, it is above
all the urban agglomerations, and especially those in the Federal States of Baden-Württemberg,
Bavaria and Hesse as well as in the city States (in eastern Germany, Dresden), that have an
above-average concentration of research-intensive branches of industry and thus an aboveaverage potential for outsourcing. Therefore, one could argue that business services still need the
exploratory advantages of the conurbation, or that within the business service sector there exist
different types of (spatial) dependencies on industry (14), or also that one conurbation can be more
(14) Brake (1994: 483) has elaborated a rule of thumb on this issue: the less routine involved and the more complex
the cooperation, the more central the locations of business services. This is particularly true for knowledgeintensive business services such as financial services and enterprise consultancy. While new information and
communications technologies in principle allow decoupling from locations, a spatial (global) distribution of
production plants, offices, branches, etc., requires centralised management functions and increases the demand
for highly specialised services for complex organisations (Sassen, 1991, 1993). The close proximity of such
services is described as an indication of a kind of just-in-time cooperation (Brake, 1993).
190
Determinants of business and personal services: Evidence from the German regions
attractive for certain business services than another (Begg, 1993). The notable preference of
multi-media companies in Germany for core cities appears to confirm this dependency (Beise et
al., 1998; on Hamburg, see Gornig et al., 1999) (15).
The advocates of the ‘global regions’ concept believe that it is primarily the regions which
have to compete internationally for globally active enterprises. While this idea contradicts the
theory of the ‘location-free’ economy, at the same time it provides further support for the
theory of polarisation or hierarchisation of spatial development (Döhl and Sauer, 1992;
Häußermann and Siebel, 1995). Regions with efficient and internationally oriented service
firms and industrial manufacturers will be among the winners in the globalisation process,
while the risk potential for regions with little human capital and with only regionally oriented
enterprises is already very large and will continue to increase (Zart and Huege, 1999).
The conclusions for political action derived from this theory are very diverse. On the one
hand, there are calls for further deregulation, which would allow the regions to adapt to the
new location requirements of globally active enterprises. On the other hand, new concepts for
structural policy in the regions, in the sense of a performance-based regional policy, are
demanded (Lammers, 1999), which, unlike regional policy to date (targeted at reducing
geographical disparities), would accept regional differences or even further reinforce them.
The alleged decline of political influence is refuted not least by the concept of innovative
networks and milieus, which refers precisely to the increased need for cooperation on the part
of the political actors. Target-oriented policies and public-private interaction is required in
particular, given the structural changes described, in the development of a long-term
infrastructure for lifelong vocational training and education (Lennartz, 1999; Reissert and
Schmid, 1999).
To sum up: The key argument to be tested is that (proper policies and favourable context
conditions given) Verdoon’s Law might work in business services and in some of the
(industrialised) personal services, whereas Baumol’s cost disease seems to hold for the rest of
the personal services. In the case of business services, it would be the ability of the region to
foster ‘service-industrial’ districts through an innovative milieu or interactive networks of key
actors in local communities and through a high level (quantitatively and qualitatively) of
training and education. In the case of personal services (which, of course, would also benefit
from a prosperous sector of business services), three alternative solutions are available in
principle: First, the allowance of large wage spreads possibly leads to more jobs in low-wage
services. However, this possibility is widely incompatible with the German constitutional
framework of liberal corporatism and social security because it leads to poverty for substantial
segments of low-skill persons. Nevertheless, even acknowledging the limits of pathdependency (the ‘German model’), there is still untapped adjustment capacity and adjustment
need in the wage formation process to which we turn briefly in the policy conclusions (see also
(15) ‘Multimedia’ is assigned the role of a leading key technology which makes the distinction between industry
and service obsolete. In the multimedia sector numerous product innovations overlap with process innovations
(for example, telework).
191
The job creation potential of the service sector in Europe
Schmid, 2000). Second, the State as employer (of the last resort) may create more service jobs.
Fiscal constraints, however, are limiting this option which in addition will only be successful
if new public employment is decentralised and organised by modern management principles
(management by objectives, accountability, etc.). The third option is an active industrial and
labour market policy through support of innovations and through training or selective wage
subsidies (or combinations of transfers and wage income). With respect to the second and third
alternatives, there is already relatively high regional discretion despite central regulations
concerning unemployment insurance and employment promotion (16).
2.3. Structure and dynamics of service employment in German regions
We start with general trends of regional employment dynamics in Germany from 1977 to
1998, and then turn to special items and puzzling items for explanation. As we already
remarked in the methodological introduction of the contribution, the internal view of German
regions and their variations might counterbalance to some extent the general view of Germany
as an ‘over-industrialised’ and ‘under-serviced’ country. Even if this might be true on a general
level, regional or location specifics (of which there are many in a federal system like
Germany) can provide insights into relationships which countervail general tendencies. This
expectation seems all the more justified under the assumption that ‘globalisation’ leads to
higher competition between regions and rewards regional innovations deviating from national
trends. We concentrate on West Germany (without Berlin) to abstract from the very special
development in east Germany; interaction of these two regions since 1989, of course, has to
be acknowledged.
2.3.1. General trends of employment
Before starting to present the results, the reader should be aware of some special features of
our data base. The figures presented in the following are based on the employment statistics
(‘Beschäftigtenstatistik’) of the Federal Employment Service (‘Bundesanstalt für Arbeit’) (17).
This data include only employees in ‘regular jobs’, namely, employees who have to pay
contributions to the social insurance system (unemployment insurance, health insurance,
senior citizen insurance). This means that employees in occasional jobs, part-time jobs under
the threshold of 15 hours, the self-employed and free professionals as well as civil servants
are not included. Thus, unfortunately, the service sector is under-represented, especially the
so-called low-wage sector. The disadvantage of this data may not be so severe if one assumes
a reasonably high correlation of ‘regular employment’ and employment not covered by the
obligatory social security system. However, the possible displacement of regular jobs by
(16) See Schmid, Speckesser and Hilbert, 2000 as well as Mosley and Schütz, 2000 who — on the basis of
aggregate impact analysis and implementation studies — found considerable regional variance of active labour
market policy even holding constant unemployment rates. They found also that training and long-term wage
subsidies contribute significantly to reduce long-term unemployment.
(17) This is a good occasion, of course, to thank the ‘Bundesanstalt für Arbeit’ for providing the data.
192
Determinants of business and personal services: Evidence from the German regions
‘irregular jobs’ (especially occasional jobs, precarious part-time jobs or feigned selfemployment) cannot be taken into account, but should be kept in mind in interpreting the data.
The data we are using are not individual data but data aggregated at the level of the 181
employment office districts in Germany. Finally, the data are not representing flows but stocks
at one point of time in the respective year, 30 June, which is the seasonally least affected
month.
(1) Our first observation is, that the overall employment performance for west Germany was
not as bad as often quoted: The loss of 1.6 million jobs primarily in industry
(manufacturing) was overcompensated by 3.8 million additional jobs in the services, a
growth rate of 44 %. Thus, Germany is on the march in the service economy; now with a
share of 59 % of service employment in 1998. What is interesting to note, however, is that
knowledge-intensive (manufacturing) industries lost in employment, too (Tables 1, 2 and
Figure 1).
(2) If we split the period into the pre-unification (1977–89) and the post-unification periods
(1989–98), we see a slightly lower growth dynamic in services in the recent period, but a
breakdown in industrial employment, especially in knowledge-intensive industries. This
indicates a failure of adjustment to ‘globalisation’ and a loss in competitiveness (Figures
2 and 3).
(3) The collapse in manufacturing was regionally unbalanced; mainly agglomerated areas
were affected and not compensated for by higher growth in service employment (Figures
4 and 5). In terms of location theory or of economic geography, the consistent pattern of
relative employment losses in agglomerated areas has to be interpreted in the way that
centripetal forces (linkages, thick markets, knowledge spillovers and other pure
externalities) have weakened, and that centrifugal forces (immobile factors, land rent and
commuting, congestion and other diseconomies) have strengthened or remained constant
(Fujita et al, 1999: 346). A first possible conclusion from this observation could be that
services in general, and dynamic services in particular rely less on local specific
externalities than manufacturing. A second conclusion is that centrifugal forces relevant
for manufacturing became stronger, for instance rising real estate costs, transport costs
and labour costs. For our research question — namely the search for favourable
conditions to create more service jobs — the specification of changing externalities
(weakening centripetal and increasing centrifugal forces) and the way some agglomerated
areas could stem this general trend, is most interesting.
(4) Although widely known, the concrete evidence with respect to the gender impact is still
astonishing: Two thirds of the new (and mostly service) jobs have been taken by women.
The overall growth rate of women’s employment during this period of 21 years (1977–98)
was 27.2 % whereas men’s employment almost stagnated (1.5 % growth). In the
agglomerated areas, the growth dynamic was considerably smaller than in the nonagglomerated areas (probably also a base-effect), men’s employment even declined.
During the last decade, however, the overall growth dynamic in services declined and
women were especially affected (Tables 3 and 4; Figures 6 and 7).
(5) In a functional breakdown of the services into five categories (business, distribution,
consumption, social and governmental), we see rocketing growth rates in social services
193
The job creation potential of the service sector in Europe
and business services as a consistent pattern over the whole period (18). Furthermore, the
dynamics of business services surpass social services during the last decade. Among the
other service branches, we observe moderate increases in distribution (retailing,
transport), consumer and government services, the latter turning into negative rates of
growth in the recent period. As expected in theory, knowledge-intensive services
contribute most to the growth in business services. If we divide services in the dual
categories of domestic trade oriented services and expert-oriented services, we see that
expert-oriented services increased a bit more than overall services but they still make up
a relatively small share of total services (1977: 20.1 %; 1998: 23.6 %); they are clearly
concentrated in agglomerated areas (1998: 27 %) (Table 1, Figures 8 and 9).
(6) The gender impact of the categorical split of services also does not come as a surprise:
about 80 % of the employment growth for women is related to domestic trade services, the
rest (20 %) to expert-oriented services; men’s proportion in expert-oriented services is
about 40 %. In terms of the five functional categories, every second woman found her job
in social services. However, women also share the same proportion of job gains in business
services, and among the business services they even gained in the knowledge-intensive
sector where they are now in the majority, especially in non-agglomerated areas. If one
accepts the share of service employment as indicator for modernisation, the agglomerated
areas are undeniably leading the development: 81 % of women now work in services
(against 75 % in non-agglomerated areas), and the corresponding figures for men are 51 %
(against 39 %). Women are now in the majority in the service sector as a whole (57 % in
1998) with a tendency towards increase, especially in the non-agglomerated areas.
However, the main factor contributing to this trend was the drastic decline of men’s
employment in government services during the last decade whereas women’s employment
continued to increase in this sector (Tables 3 and 4). Within our five functional categories
of services, women are now under-represented only in the distributive services. In the dual
categorical split, however, women are still in the minority (44 %) but gaining considerably
in the expert-oriented services, and in the majority in domestic trade oriented services
(61 %) (Tables 5, 6 and 7; Figures 10 and 11). A tentative interpretation of these patterns is
that women’s over-proportional gain in service employment is related first to their
traditional proximity to the rising service activities, second to their relatively lower wages
compared to men, and third to their greater willingness to accept flexible employment
relationships related to the new jobs, for instance part-time or fixed-term contracts.
(7) Apart from the generally lower growth dynamics in services of agglomerated areas, the
regional pattern is relatively balanced. Most noticeable is the lower capacity of
agglomerated areas to create simple service jobs which is indicated, for instance, by the
considerably lower growth rate of domestic trade oriented services compared to nonagglomerated areas. We assume, however, that this reflects the lack of marginal jobs
(‘630-mark jobs’; see Appendix 1) which are not counted in the statistics used here and
which are very likely concentrated in agglomerated areas. In addition, ‘black work’ or the
(18) For a specification of these aggregated branches see Appendix 2.
194
Determinants of business and personal services: Evidence from the German regions
illegal informal sector is growing in Germany and probably concentrated in the cities.
Thus, the observed pattern of relatively low employment dynamics in agglomerated areas
can to some extent be interpreted as a substitution of regular employment relationships
through informal or even illegal employment.
(8) If we go a bit further into the details concerning the knowledge-intensive sectors, we see
heavy losses in the machine tool and the electric and electronic industries (Germany’s
pillars of manufacturing), and within the services we observe failures in financing and
insurance jobs but also a lack of dynamics in the engineering services. The foundation of
Germany’s knowledge-intensive business services consists of counselling, marketing and
advertising services. Expert-oriented services increased a bit more than overall services
but they still make up a relatively small share of total services (1977: 20.1 %; 1998:
23.6 %); they are clearly concentrated to agglomerated areas (1998: 27 %).
2.3.2. The influence of sectoral branches on regional employment
In the next step we look at some simple linear correlations on the basis of the 12 largest
agglomerated areas (taking the non-agglomerated areas as a contrasting case) (19). The aim is
to identify interesting patterns as first hints of the determinants of the regional employment
dynamics.
(9) There is a long-standing debate on the linkage between manufacturing and services. Even
if manufacturing employment (here ‘industry’) is declining due to productivity effects,
according to the ‘base-multiplier analysis’ (‘export-base theory’), a competitive core of
expert-oriented industries is crucial for a sustainable regional economy. The export
industries or services multiply employment because many services or intermediate
industrial goods are required as inputs to the manufacturing industry. In the extended
version of interaction theory we have argued that high quality services are required to
maintain competitiveness in the industry. This linkage or complementarity hypothesis is
confirmed by the strong correlation between industry and total employment (Figures 12
and 13) and between industry and services in particular (Figures 14 and 15) However, the
correlation deteriorates and supports the argument that the regional service economy
becomes more and more independent from industrial development. At this level of
observation it cannot be determined whether export-oriented services replace exportoriented industries, or whether heavy outsourcing and corresponding downsizing of
industrial production lies behind the eroding or even the negative correlation between
industry and services most recently.
(10) The interaction theory also seems, at first glance, to be refuted in the recent period.
During the 1980s, business services and industry were still positively correlated, but the
relationship reversed during the 1990s and became negative, with more supporting the
(19) These areas have been composed by the aggregation of employment office districts (‘Arbeitsamtsbezirke’)
which build functionally integrated labour markets; for details of this procedure see Albrecht/Schmid (1985).
For the specification of these functionally integrated labour markets (agglomerated areas) see Appendix 3.
195
The job creation potential of the service sector in Europe
outsourcing theory or the theory of relative autonomy of services (Figures 16 and 17).
We have to be careful, however, with this interpretation because some of the breakdown
in the relationship may be due to the impact of the 1992–93 recession which heavily
affected manufacturing whereas services seem to be less sensitive to business cycles.
Nonetheless, there are indications that globalisation and information or communication
technology weaken the spatial ties between manufacturing production and services,
especially business services which interact more and more globally.
(11) On the other hand, the extended interaction theory is clearly confirmed by the fact that
the overall employment dynamics depend heavily on the existence of the combined
knowledge-intensive services and knowledge-intensive industries in the region. The
correlation is quite strong (1977–89: r = 0.62) and becomes even stronger (1989–98:
r = 0.70) (Figures 18 and 19).
(12) The globalisation (or de-linkage) theory is also confirmed by the fact that all other
correlations between specific branches and overall employment dynamics become much
weaker in the recent period. For instance, the strong correlation between expert-oriented
services and overall employment growth r = 0.74 (1977–89) fell to a level of r = 0.10
(1989–98).
2.3.3. Structural and location determinants of service employment
A shift-share-analysis can reveal to what extent a favourable regional development is due to
‘structural’ or to ‘location’ determinants (20). ‘Structural’ determinants reflect the composition of
branches corresponding to the overall development of the economy, and ‘location’ determinants
hint at specific characteristics of the region. These specifics can be changes in the ‘geostrategical location’ of the region (for example, belonging to the ‘European group’) or, and of
most interest for us, in institutional arrangements or explicit policies. A positive structural
component of growth indicates that the region was favoured by the overall service development,
whereas a negative structural component of growth indicates an unfavourable mix of services.
A positive location component of growth reflects some hidden institutional or geo-political
factors and just good luck, and a negative location component indicates institutional or political
failure or just bad luck. We now turn to the results of the shift-share analysis on the basis of
agglomerated areas contrasted to the (one case) non-agglomerated areas (see Tables 8, 9 and 10).
(13) Although agglomerated areas were favoured by the mix of services (structural effect),
their strong negative location effect hints at the main reasons for the low or modest
employment performance. The hypothetical employment growth of 46.9 % (between
1977–98) was only reached by 82 % (38.7 % realised growth rate); such a strong
negative location effect indicates adjustment failures or bad luck. On the other hand, the
non-agglomerated areas could compensate for their lower structural advantage to some
extent by location specifics.
(20) The shift-share analysis refers only to service employment; the same analysis with overall employment
reached almost the same results with some changes in the ranking of regions.
196
Determinants of business and personal services: Evidence from the German regions
(14) The breakdown of the period shows that the negative locational factor for the
agglomerated regions remained constant but the positive structural effect declined
drastically. In other words, the recent relative decline in the employment performance of
the agglomerated areas is mainly due to structural effects and not to a worsening of
locational effects.
(15) The figures, however, show drastic differences in the locational effect. Some cities (and
their immediate surroundings) have positive locational effects (Nuremberg and
Rhein/Neckar) during the 1980s, and some cities have consistently high negative effects,
especially Hamburg and Bremen. The differences in the rankings (1977–89, 1989–98)
also indicate interesting changes: Cologne and Hanover performed relatively better in
the 1990s than in the 1980s when Munich and Stuttgart seemed to be model regions.
Nuremberg is a consistently high performer. We will return to speculate about the hidden
‘institutional’ factors or possible changes in the ‘geo-political arena’ later (21).
(16) The much stronger correlations between the locational effect and total effect (equal to
total employment growth) compared to the correlations between structural effect and
total effect indicate that there is more homogeneity between the agglomerated areas with
respect to the structural component than with respect to the locational factors (Figures
20, 21, 22 and 23). In other words, to gain further insight into the determinants of
regional employment performance, we must look more closely at the locational rather
than the structural effects.
2.3.4. Employment rates and sectoral mix
One way to approach the difficult question of the ‘hidden determinants’ behind the locational
effect, is to look at the relationship between employment rates and changing sectoral mix (22).
The employment rate here is defined as the number of actual employed persons in the region
as a percent of the total population (residents) in the respective region. This indicator, thus,
does not measure labour force participation but rather the workplace intensity of a region
relative to the number of its population. Nevertheless, it can be taken as a measure of
performance indicating both regional attractiveness and social integration through gainful
employment (Schmid, Schütz and Speckesser, 2000). The employment rates vary between
26.7 and 43.5 % (mean = 36.2 %) in the agglomerated regions. Due to the attractiveness for
commuters from ‘rural zones’, these rates are higher in agglomerated areas than in the nonagglomerated areas, although the difference diminished during the period of observation. In
the following, we compare the three regions with the highest employment rates (Düsseldorf,
München, Stuttgart), and the two regions with the lowest (Essen and Bremen).
(17) The first observation shows some commonalties between high performers but also
different sectoral mixes especially related to industries (the ‘secondary’ sector). Stuttgart
is a case with continued high rates of manufacturing whereas the high employment rate
(21) About the ‘model Baden-Wurttemberg’ compared with the region of Massachusetts, see Sabel et al. (1991).
(22) We intend to go deeper into this question in follow ups of this report.
197
The job creation potential of the service sector in Europe
of Düsseldorf and especially Munich can be mainly traced to high rates in services.
However, it is important to note that even Düsseldorf and Munich still have employment
shares in manufacturing industries which are not far below the national level. The
commonalties relate to knowledge-intensive business services and domestic-trade
oriented services, whereas other business services, distribution and export-oriented
services (high in Düsseldorf) and consumer or social services (especially high in
Munich) vary considerably among the regions (Table 11 and Figures 24 and 25).
(18) The low performing regions of Essen and Bremen were traditional mono-centred areas
(shipbuilding and fishing in Bremen, mining and steel industry in Essen) (23). The
breakdown of these industries could not be compensated for by modern industries or
business services. Thus, both regions have not only a tremendously low rate of industrial
employment but are also below average in modern services, as well as meeting only
average standards in social and household related services (Table 12 and Figures 26–27).
(19) Some first-hand statistical relationships provide hints for explaining the high or low
performance of regions. We find a significant negative correlation between employment
rate and share of social services (r = – 0.45), and a strong positive correlation between
employment rate and share of business services (r = 0.65) as well as a strong correlation
between employment rate and the combined knowledge-intensive industries and
services (r = 0.56); the relationship between employment rate and knowledge-intensive
business services is even stronger (r = 0.70). In all three cases, the three selected high
performers have top scores whereas the two low performers (Bremen and Essen) fall
even under the expected values of the regression line (Figures 28–30).
To sum up this section: High performance in terms of the employment rate can be connected
with at least two production clusters. The agglomerated area of Stuttgart is still characterised
by a dynamic manufacturing sector connected with knowledge-intensive industries and (to a
lesser extent) with knowledge-intensive business services. The other two high performers
(Munich and Düsseldorf) are focused on services but still have a substantial and polyvalent
industrial core. Within services, there are clearly knowledge-intensive business services
which are the driving motor of the employment dynamics, and knowledge-intensive industries
are also strong. Both service sectors are not well developed in the low performing areas of
Bremen and Essen. This can be interpreted as a strong indication for the overall importance of
knowledge for sustainable regional employment dynamics. In other words, new service jobs
are created especially in regions in which knowledge-intensive services and knowledgeintensive industries interact. Especially in Munich and Stuttgart (with Nuremberg and
Frankfurt close behind), we find the characteristic elements of ‘innovative clusters’ (Simmie
and Sennett, 1999): Of the seven knowledge-intensive industries, five were over-represented
in Munich, and four in Stuttgart; of the four knowledge-intensive business services, four were
(23) The agglomerated area Essen (one of the largest cities in the ‘Ruhrgebiet’) is an example of the decline of a
once prosperous industrialised area. These — in a life-cycle analogy — ‘old’ industrialised areas are the
‘victims of an early form of division of labour’ (Steiner, 1985: 396).
198
Determinants of business and personal services: Evidence from the German regions
over-represented in Stuttgart, three in Munich, and two in Düsseldorf (with an exceptionally
high share of public relations and advertisement) (Tables 13 and 14).
2.4. Summary discussion of the results and policy conclusions
The following final section summarises and discusses the main results of the empirical study
in the light of the theoretical framework (2.4.1). Since the service gap in Germany is mainly
related to the so-called low-skill sector of personal and domestic services (at least in public
opinion from inside and outside Germany), we take up this discussion in more detail (2.4.2),
provide our own view on this issue (2.4.3) and finish with policy recommendations (2.4.4).
2.4.1. Evidence on business and personal services from Germany
The most important result related to business services is the interactive nature of the
knowledge-intensive sector within business services and knowledge-intensive
(manufacturing) industries. Whereas the overall nexus between industries and services
declines, this nexus is the crucial determinant for explaining positive and sustainable
employment dynamics. This means that the manufacturing of products (the so-called
secondary sector) remains crucial, but these products contain more and more service elements
or are closely connected with them (‘hybrid products’). A corollary feature is the strong
correlation between the skill level of the regional labour force and the regional employment
growth, especially related to export-oriented services and business services.
In terms of the spatial impact we find that high-skill and expert-oriented services and
knowledge-intensive industries are still concentrated in agglomerated areas. However, there
are indications that they move more and more to the periphery of these agglomerations. Lowskill services seem to locate in the centre of agglomerated areas (24) or at the rural periphery.
On the other hand, large-scale internal labour markets in manufacturing, primarily in
agglomerated areas, are transformed into network labour markets. The headquarters and skillintensive manufacturing parts of those industries may remain in the agglomerations but semiskilled and low-skilled production is moving to low-wage regions or to the proximity of the
foreign sales markets. This is why especially agglomerated areas were confronted with
dramatic job losses in industries. Only those regions could compensate for this loss by being
able to establish interactive networks of knowledge-intensive services and industries.
The interaction between knowledge-intensive services and industries also has spillovers into
other domains. Increasing demand in service intensive manufacturing goods leads to a parallel
development (positive correlation) of employment in industries and services; this creates
higher income thereby higher demand for personal services thus creating a virtuous circle (as
well as a vicious circle in regions in which the industrial core loses its foothold). Interactive
(24) Which we cannot test with the kind of data we have; see the box on ‘630-mark jobs’ in Appendix 1.
199
The job creation potential of the service sector in Europe
services may also reverse the conventional product cycle: intermediate services turn into
(exportable) service goods such as software packages, management and expert systems.
Export services still require personal interaction, that is, personal presence on the foreign
markets, thereby language and communication capabilities. The culture of mobility is crucial
for competitive advantages in high-skill services; this is a further explanation for the strong
correlation between skill level and expert-oriented services.
There is a strong correlation between high-cost industry and high-cost services; this might, as
a paradox, improve the competitiveness of export manufacturing (‘made in Germany’), but
diminish the competitiveness of (high skill) services which require a low-cost regime due to
international competition, for instance by cheap computer specialists in Poland or India. As
far as this relationship works, the correlation between manufacturing and knowledgeintensive services is eroding; we found some indication of this.
Two of the most important determinants of successful adjustment to globalisation are
information and communication networks. They are the crucial lever to increase productivity
in business services and knowledge-intensive industries with likely spillover to personal
services. The reason is that information networks increase their efficiency with rising numbers
of participants, probably exponentially. The existence and public support of such networks
explains regional differences. We were not able to test this assumption explicitly; however, the
strong correlation of ‘locational factors’ (‘Standortfaktoren’) with the overall employment
dynamics can be taken as an indicator; and anecdotal evidence from case study material
confirms this expectation. Thus, taking the main results relating to business services together,
there is justification to extend the ‘industrial-district hypothesis’ to the ‘service-industrydistrict hypothesis’.
As far as personal services are concerned, the main result is the confirmation that the road in
the service society is the road of women into the system of gainful labour market work. This
road of professionalisation changes the form in which the female labour potential is organised
but hardly the contents; thus, regions with high personal or social services have a higher
female labour force participation and vice versa; this relationship is especially pronounced in
the large cities, namely, in the agglomerated areas. A corollary to this finding is that demand
for commercialised and professional personal services rises with qualification, especially with
the skill level of women. Thus, we find higher service employment rates in regions with high
skill levels, a pattern that correlates with agglomeration as the skill and income level in these
regions is higher than in rural areas (25).
Wage subsidies for low-wage personal services combined with social innovations such as
service centres or services agencies induce sustainable social service employment in regions.
Regions with high expenditures for public work creation or wage subsidies have, ceteris
(25) Especially regarding these aspects, follow-up studies will have to complement the present evidence with more
precise information for which data are lacking.
200
Determinants of business and personal services: Evidence from the German regions
paribus, higher employment rates and higher increases in household-related social services
than regions with lower expenditures (Figure 31) (26).
The spatial impact of personal services can be hypothesised in the following way: There are
two models of organising many of these services (see also Section 2.4.3): First by
industrialising service functions plus self-servicing (the ‘inward-shift thesis’, for example,
Gershuny 1983), second by large wage differentials that make buying services affordable (the
‘outward-shift thesis’, for example, Esping-Andersen 1999). The industrialisation process of
services substitutes high-cost services by highly sophisticated durable goods combined with
self-service (car driving, using washing machines, videos, personal computers, home trainers,
sauna equipment, haute cuisine facilities); this kind of self-servicing requires an increase in
household income through both rising labour force participation of women and time. The
alternative model would be to use cheap personal services (taxi drivers or public transport
personnel, laundries, cinemas, consumer, counselling, tax advising, banking services, fitness
centres and restaurants); this service model requires less free time but an even higher
household income through full- time work to afford buying the services (27).
2.4.2. Curing the service gap by wages or consumer subsidies
In theory, the high employment effectiveness of the elementary services sector (especially for
the low-skilled) is often explained by the fact that such services usually require little
knowledge or capital (Kommission, 1997), as well as by being geared towards the domestic
market, they are not subject to productivity pressure and foreign competition (as they usually
offer locally delivered and locally consumed services) (Zukunftskommission, 1998). In a way,
this line of argument follows up the ‘three-sector theory’ postulated by Fourastié (1969).
While the contrasting theory of ‘cost disease’ (Baumol, 1967) is increasingly being called into
question in connection with the spread of new information and communication technologies
and in particular with respect to enterprise-related services, a cost reduction via increased
productivity still appears out of the question for the majority of household related services (28).
Support for the Anglo-Saxon model of a wider wage spread, as recommended by the
Commission of the Free States of Bavaria and Saxony (Kommission, 1997), remains the
exception to date in the German debate, with reference being made to the high social followup costs (Semmler, 1998; Werner, 1997). Likewise, the Scandinavian model is to be excluded
as a possible solution in view of the anticipated political opposition to the tax implications
(Scharpf, 1998). There is, therefore, an increased debate recently about a middle road between
the two possibilities, namely the creation of a formal (that is, State-regulated and Statesubsidised) low-wage sector. Two possibilities for an expansion of the legal demand for
(26) See Spee and Schmid, 1995; to be tested for recent periods in follow-up studies.
(27) The expectation, that the second model is more common in agglomerated areas than in non-agglomerated areas
would be confirmed by the lower rate of part-time work of women (if controlled for the sectoral structure), and
by the higher rate of consumer and distributional services in agglomerated areas.
(28) See Tronti et al., 2000; Albach, 1989; Baumol et al., 1985; Schettkat, 1996.
201
The job creation potential of the service sector in Europe
elementary services are being discussed: reduction of wage costs by means of employee wage
subsidies and/or public support for consumption via consumer subsidies (29).
The debate focuses on three types of employee wage subsidies. First, the combination wage
model (‘Kombilohn’) (cf. Bäcker and Hanesch, 1997; Becker, 1998; Dreger et al., 1998;
Gerster and Deubel, 1999; Kaltenborn, 1998; Vierling, 1998) provides for a reduction in
labour taxes for recipients of a welfare benefit who have extra earnings from employment.
Second, under the income allowance model (Scharpf 1993, 1994), gross hourly wages of up
to DEM 15 would be increased by degressive subsidies. Third, the reduced social
contributions model (Schreiner, 1997; Zukunftskommission, 1998) represents a combination
of increased net incomes for workers and reduced gross wage costs for employers: up to an
hourly wage of DEM 10, compulsory social contributions are borne by the State alone; the
subsidy is then degressively reduced (up to an hourly wage of DEM 18).
The range of the three models (that is, the target group reach and the related volume of the
subsidy), varies considerably. Scharpf’s model has the widest range because it does not
involve a means test; thus, the household situation of the beneficiary has no influence on the
rate of income allowance. The combination wage model has the shortest range because
beneficiaries must already be in receipt of a welfare benefit; at the same time, the basis of
calculation is determined by way of a means test (unemployment assistance, social assistance)
and the amount of the tax-financed additional income granted thus depends more on the
household situation. Different estimation procedures are used to determine additional
employment effects. Employment effects can be derived from potential values (30) or by using
model calculations based on demand and supply elasticities. The latter approach yields
considerably lower employment effects (31).
Another proposed strategy of stimulating job creation in low-skill services is the support of
effective demand from private households for such services through consumer subsidies. In
Germany, this option is represented by the tax model introduced in 1989 (32). The employment
(29) The third possibility is wage subsidies for employers. Because this form of wage-cost subsidy is not
specifically relevant to the question of the low-wage sector, it will not be detailed here. In contrast to wage
subsidies for employers, employee wage subsidies would be granted, as a rule, for an unlimited period and,
with the exception of the combination wage model, the restriction of the subsidy, that it is for only difficultto-place unemployed would be lifted. Spermann (1999) points out that wage subsidies for employers (cf.
Snower 1997) would yield lower employment effects because of the associated higher transaction costs and
stigmatisation effects.
(30) Bellmann (1998) and Seewald (1999) calculate the workforce potential that can be tapped among recipients of
social assistance through combination wage models as being around one million, or two thirds of the total
beneficiaries.
(31) Kaltenborn (1998) and ZEW (1998) arrived at additional employment effects of approximately 17 000 to
18 000 persons on the basis of the combination wage model. The corresponding labour market effect for the
State of Saxony-Anhalt has been estimated at 2.4 % of all social assistance beneficiaries (cf. Dreger et al.,
1998). As regards the model involving a degressive subsidy towards social insurance contributions, Bender et
al. (1999) have calculated additional employment amounting to a total of 140 000 persons. The cost of each
additional job would amount to around DEM 80 000 per annum (cf. Schupp. et al,. 1999).
(32) Under this model, families with at least two children (at least one aged under 10), single parents with at least
one child aged under ten and families with persons requiring care are entitled to a tax deduction of up to
DEM 12 000 for the costs of domestic personnel.
202
Determinants of business and personal services: Evidence from the German regions
effect expected on the part of policy-makers was 100 000 insured jobs in the area of private
households (cf. Munz 1996) (33). In 1997 the tax deduction was raised to DEM 18 000, the
restrictions on eligibility were lifted and a so-called ‘household chequebook’ (service cheque)
was introduced. This type of service cheque is a more restrictive option (it ultimately only
certifies to the finance authorities the entitlement to claim a tax benefit), than the servicecheque or coupon models (Finger, 1997). The employment effects policy-makers hoped for
fluctuated between 500 000 and 600 000 persons (cf. Müller-Siebrands, 1999). By the end of
1998, however, only 8 449 household cheques had been used (cf. Sommer, 1999) (34).
The weak employment effect to date of this tax model is primarily traced back to this
instrument being only advantageous for households with a very high income and in
considerable need of assistance. This criticism was the departure point for the transfer model
for service cheques or coupons proposed by the German Social Democratic Party (Schreiner,
1997) (35). The service coupons are to be used to pay for corresponding services from service
agencies; private households are not liable for any employer obligations. The expected
employment effects range from 330 000 (Finger, 1997) to 700 000 full-time jobs (Schreiner,
1997). Schettkat and Fuchs estimate an employment effect in Germany of around 170 000
full-time jobs on the basis of 50 % co-financing of care of invalids (aged 70 and over) via
service cheques, while the same financing model would result in the creation of 214 000 fulltime jobs in the area of childcare (cited in Europäische Kommission, 1995).
Care insurance (‘Pflegeversicherung’) introduced in Germany in 1994 can also be interpreted
as a consumer subsidy model in the area of care services. Under this model, a maximum of
DEM 2 800 per month for material care services (payment of outpatient services) or a
maximum of DEM 1 300 per month care allowance (compensation for related extra
expenditure on the part of the person providing care) is made available. Although up to
150 000 additional jobs were expected in outpatient care, the actual employment growth in
this area was around 70 000, including 20 000 low-skilled workers (Pabst, 1999). The
difference between the predicted and actual employment effect is because 84 % of the total
1.2 million persons requiring care preferred the care allowance to material services. The
option to choose was certainly politically desired in the sense that it represented recognition
of and pension insurance for care provided within the family (Evers, 1995); it was also availed
of, despite the notable difference to the monetary value of the material services.
(33) In 1989 the number of insured employees in the area of private households was 32 000; this figure had risen
to 34 000 by 1997 (cf. Bittner and Weinkopf, 1999).
(34) Finger (1997) estimated the employment effect of this model at 14 000 full-time jobs. Munz (1996) arrived at
an employment effect of 130 000 (on condition that the tax deduction be raised to DEM 24 000) or 180 000 if
there were no ceiling on the deductible amount.
(35) Under this model, households (with at least one child aged under 14 or one person requiring care and aged at
least 80) receive service cheques to the value of DEM 1 200 per annum (and an additional DEM 600 per
additional person).
203
The job creation potential of the service sector in Europe
2.4.3. Evaluation of the present discussion
The discussion presented so far has pointed out the substantial discrepancy between estimated
employment potentials in low-wage or personal service jobs and actual expansion to date.
Before turning to policy proposals, some reflections on the causes of this discrepancy are
required. Contrary to the thesis of an outward shift of production, a range of authors supports
the thesis of an increasing inward shift of production into private households. Falling output
prices and rising wage rates for working women favour the substitution of time-intensive
household production by capital-intensive household technologies (Ott, 1998). However,
Meyer (1997) maintains that this substitution — in conjunction with a tax system that favours
the ‘wife-at-home’ model — also reinforces the trend towards a self-service economy, which
in turn makes the idea of delegating household services to third parties seem unattractive.
Schweitzer (1995) believes that the preconditions for the welfare-based shift of production
into private households will be increasingly called into question in the future in view of the
erosion of the standard employment relationship based on the ‘breadwinner model’. The trend
towards a shift of production into households, however, might be reinforced through the rise
of precarious jobs and long-term unemployment, while home or individual work would
become more significant as a source of livelihood.
Another obstacle to the establishment of a formal low-wage sector is the existence of an
informal low-wage sector, estimated for Germany at around 1.4 million employees in the area
of private households (Ochs, 1999) and confirming corresponding job creation potentials in
this domain calculated by various authors (cf. Albach, 1989); 2.8 million households regularly
employ domestic help, while 1.5 million households do so periodically (Schupp. et al., 1997).
In other words, more than every tenth private household in Germany use such services (36). At
present, the formal low-wage sector is unable to compete, either on the supply or the demand
side, with the informal sector because of the prevailing price discrepancies (37). If the
‘elementary activities’ sector is discussed only as an alternative to public services in this area,
it can be implied that an insufficient supply of public services (especially in childcare) is often
the basis for the establishment of an informal low-wage sector, which in turn becomes a
limiting factor for the regulated low-wage sector (38).
On the other hand, if one looks at the sector solely as an alternative to the informal
employment sector, the only strategy for establishing a formal sector is the path of price
competition via publicly subsidised low prices or a reduction of the legally guaranteed
(36) To compare: there are currently 62 service agencies in Germany with 1 047 employees and 3 571 customers.
Three quarters of these agencies have received or are receiving publicly funded project support, while 95 %
have made use of wage-cost subsidies (cf. Zukunft im Zentrum, 1999a).
(37) The average (contribution-free) hourly wage paid in the informal service sector is between DEM 15 and
DEM 25 (Odierna, 1995). Service agencies charge their customers an average of DEM 23 an hour, while the
gross wage paid to their employees is an average DEM 15 an hour (cf. Zukunft im Zentrum, 1999a). This
discrepancy results, among other things, from paid holiday leave for the employees, continued pay in the event
of illness and provision of training measures.
(38) Consequently, the co-existence of a trend towards a shift of services into the household and an increasing
supply of services on the ‘grey’ labour market is not necessarily a contradiction.
204
Determinants of business and personal services: Evidence from the German regions
minimum income. The latter path is more ‘economical’, but it is not certain whether a
corresponding demand would follow, as demand elasticity in this area is not only determined
by prices but equally by the quality and the ‘image’ of the services (Zukunft im Zentrum,
1999b). The removal of these qualitative reservations is linked in turn to the recruitment of
suitable employees, which is prevented by the low wage incentives.
At the same time, the problem of a qualitative service gap, said to limit the quantitative
demand for services, is also noticeable. New and qualified fields of employment are said to be
found in the integration of care and supervision services with housekeeping and maintenance
work on domestic technology and the home (‘associated services’), in the interface between
new information and communication technologies and private households (home information
technologies), in advisory services, and in the areas of culture and leisure (Bopp, 1996; Meyer
et al., 1999; Mundorf, 1999; SOFI, 1999). This path would also be more in line with the
opinion that Germany’s employment problems are not primarily caused by a gap in
elementary services, but are rather the result of a gap in high-quality, knowledge-intensive and
enterprise-related services (Gries and Birk, 1999; Semmler, 1998; Werner, 1997; Freeman and
Schettkat, 1999). The definition and elaboration of these new fields of employment will be an
important task for further research in the area of household related services.
The insufficient exploitation of the employment potential of household related services is
generally recognised. What remains controversial are the ways and means of tapping this
potential. The criticisms made as regards the creation or expansion of the low-wage sector are
the following:
— Substitution effects between ‘regular’ and subsidised jobs and resulting deadweight
effects can be expected (Bäcker and Hanesch, 1997; Becker, 1998; Bofinger and
Fasshauer, 1998; ZEW, 1998), while the incentive to invest in human capital would
decline (Schettkat, 1996).
— Because these services are made use of largely by high-income households, public
subsidisation of such employment relationships would have a regressive distribution
effect (SOFI, 1999).
— As regards the effect on the level of unemployment, it should also be noted that these
measures would probably increase the labour market participation of persons who had
previously not sought work (Bender et al., 1999; Schupp. et al., 1999).
Despite these objections, some authors also recognise the fact that a wider wage spread would
increase the demand for household related services and at the same time improve the
employment prospects of less competitive male and female workers. A wider spread of gross
wages could be compensated for by easing the contributions burden via increased funding of
the social insurance system from tax revenue (Bender et al., 1999; Bofinger and Fasshauer
1998; Walwei, 1998). We find this argument convincing in as far as the general contribution
level to social security could be reduced and if some current expenditures not directly related
to social insurance (such as for adult training for German immigrants from eastern Europe)
would be financed from the general State budget. There is surprisingly little discussion about
support for the supply side by means of a reduction in or exemption from liability for VAT
205
The job creation potential of the service sector in Europe
(Zukunft im Zentrum, 1999a), although this would entail much less intervention into the
taxation, contributions and wage systems. The Council of the European Union has also
recently suggested, as a step towards more employment-friendly taxation systems, that the
VAT rate should be reduced for labour-intensive services (Europäische Kommission, 1999).
Our study predicts further support for this proposal: To the extent that a favourable tax
structure enhances the professionalisation of household related services, prices will fall in the
long run and psychological barriers of buying such services will decline (39).
2.4.4. Policy recommendations
Our study suggests two pathways to increase the job potential in services: improving the
competitiveness of high quality and expert-oriented business services, and relieving the cost
burden for personal or labour intensive services. With regard to the first pathway,
competitiveness of business services can mainly be improved through innovation and
increased productivity. The most promising policy orientations that correspond to the results
of our study are the cultivation of innovative milieus or networks at the regional level.
However, the concentration on business services is too narrow. A related but often neglected
path for fostering services is to have specialised high quality and high skill personal services
in education, health, culture and tourism. Through the digital revolution (information and
communication technology), these personal services become more and more tradable or can
attract ‘service tourism’ (foreign students, clients for special health or body treatments,
consumers of cultural events, etc.). In addition, as our study has confirmed, product
innovation in manufacturing is still an option and a promising orientation for industrial policy,
for example, solar-energy or bio-technology.
The effective governance of networks is still not well understood. However, four lessons
appear to have achieved widespread consensus in reviews of network literature: First, most
effective regional networks are voluntary private enterprise networks in which policy-makers
have a very limited role. The most important support of such networks is indirect through
provision of ‘hardware’ infrastructure (for example, excellent transport and communication
networks) and ‘software’ infrastructure (for example, a skilled and competent labour force,
information clearing). Second, some effective regional networks are public-privatepartnerships in which regional governments or municipalities enter as co-financiers and,
therefore, as important bargaining partners. City contracts, regional pacts or target specific
associations (in which local or regional governments are members) are common elements in
successful public-private-partnerships. Third, policy-makers can play an important role as
(39) A model project in Berlin turned out quite successfully: The agency ‘Inhouse’ (Privatinstitut hogan) trains and
employs personal managers, care managers, event managers and ‘butlers’ who offer professional services to
private households in terms of care for children, the disabled or elderly; housekeeping, maintenance or
administration; organising festivities; private secretarial work; garden work, etc. The agency is now
independent from public support. It employs 16 persons on a permanent basis or on the basis of a guaranteed
monthly honorarium, paying an hourly salary between DEM 23 and 30 (on average higher than informal
services of this kind); for further information see Hogan 1999. Further arguments for reducing or
differentiating VAT in favour of services are provided by the contribution of Schmid in this volume.
206
Determinants of business and personal services: Evidence from the German regions
‘gatekeepers’ or moderators in the formation of regional networks. This follows from the
reflexive use and transmission of knowledge as the core of an information or knowledge
society. Knowledge of knowledge requires translators or interpreters of information. In
playing the role of ‘moderators’, local or regional governments take over an entrepreneurial
role which indirectly may lead to new jobs in new service markets. Fourth, networks are no
panacea. Conventional strategies of economic policy (setting proper market frameworks and
compensating obvious market failures), and social policy (setting proper frameworks of social
security and equal opportunity policies) are still required.
The second path for fostering personnel or labour-intensive services is the extension of basic
income support to persons with low earnings capacities and the professionalisation of
household-related services. Several routes are open to stimulate (low skill) personal services.
However, the general strategy is clear: a widening of the wage spread combined with various
methods of wage subsidies, income policy or tax differentiation. The route to be selected
should be left to the individual employment regimes in the European Union’s Member States
taking into consideration path dependency and slow learning capacities of institutional
routines. In the German context, the start-up support of local agencies providing home-related
services, the employment-friendly differentiation of non-wage costs (especially employers’
contributions to social security), and a differentiation of VAT are probably most promising in
terms of effectiveness and political feasibility.
207
The job creation potential of the service sector in Europe
Appendix
Appendix 1: Marginal jobs
Box A1: Service jobs outside the regular labour market in Germany (‘630-mark-jobs’)
This form of contract, valid until the end of 1998, meant that the employee was, by law,
registered by the employer, who had to report all details to the chamber of trade, but the
employee was only allowed to work for less than DEM 630 a month (EUR 321) and not for
more than 19 hours a week. These were not ‘black’ jobs. They were legal but did not count
in the employment statistics because the persons were not regular employees. To some
extent this was and still is a special German variation of low wage jobs excluded from the
social insurance. No payments were made to the obligatory social insurance, no tax was
paid by the employee. The employer paid a low ‘head tax’ for each contract. The offers of
the public employment service (PES) could be used as these persons were not registered as
employed with full rights and duties, nor as unemployed.
These particular job facilities were originally meant for students and pensioners, but were
not restricted to these groups. So there has been a steady increase of these jobs in many
areas such as restaurants, hospitals, air bases, retail shops and so on. Between 1992 and
1997 a rise of 80 % of this type of contract was registered. It is easy to see why. Employers
could avoid paying high labour costs, the social costs and could neglect other rights the
employees have in regular contracts. In 1997, a total of 5.5 million such contracts were
signed (about 15 % of the total active labour force). The women made up 90 % of these.
Twenty-eight percent of those with such contracts worked in households, 18 % in retail
sales, 15 % in industrial trades, 10 % in the restaurant and business cleaning services.
However, such contracts were also used in teaching institutions (for example, in the
Volkshochschulen or private schools), or as personnel attached to private medical doctors
and dentists, to welfare organisations, newspaper delivery, or to private hospitals. In the
households of 1997, a total of 1.4 million such domestic jobs were registered. This does not
include those who work privately without any registered contract at all; 500 000 to 1 million
households are estimated to employ persons in such a way, without any formal regulations
or a contract. The agreement is totally private on both sides, but it also hints at the shadowy
existence of a great many household services (Fölster, 1999).
The new government in power since October, 1998 has, as a first step, turned the tax
obligation of the employer into social security contributions (12 % to the pension system,
10 % to health insurance); in order to receive full entitlements from the pension system,
employees must contribute an additional 7.5 %. The reason for this (hotly contested)
decision was to avoid the increasing substitution of full-time work (underlying social
security contributions) by ‘630-mark’ jobs; non-regular 630-mark jobs for up to two months
or 50 working days per year (seasonal jobs) are still completely exempt from social security
contributions. The impact of these legal changes is not yet clear. First hand evidence,
however, shows that neither the fears (dramatic decline of these jobs) nor the hopes (drastic
increase of regular low-wage jobs) were realised.
208
Determinants of business and personal services: Evidence from the German regions
Appendix 2: Classification of industries and services
Knowledge-intensive industries/
business services
Knowledge-intensive industries (1)
Knowledge-intensive business
services (2)
Other business services
Distribution services
Consumer services
Social services
Government services
Domestic-trade-oriented services
Expert-oriented services
Household related services
Industries/services with a large concentration of
professional occupations; professional occupations include
such jobs as engineers, computer scientists, natural
scientists, physician, lawyers and architects.
Chemicals and allied products, machinery, motor vehicles
and equipment, aircraft and parts, electrical machinery and
equipment.
Financial services, legal services, management services,
engineering and architectural services, research and testing
services, public relations services and advertising.
Services to buildings, rental services, business
associations, mailing, reproduction and stenographic
services and miscellaneous business services.
Wholesale and retail trade, transportation and
communication.
Hotels and other lodging places, restaurants, beauty shops,
entertainment, media, publishing houses, photographic
studios, individual and family services.
Childcare and senior services, educational services, health
services, civic and social associations.
Public administration and social insurance.
Other business services, retail trade, consumer, social and
government services.
Knowledge-intensive business services, transportation and
communication.
Wholesale and retail trade, consumer services, childcare
and senior services.
(1) Total employment share of engineering and science technician occupations in manufacturing in West
Germany (1995): 2.7 %; chemical industry: 6 %; machinery: 6.1 %; motor vehicles: 3.5 %; aircraft: 16.3 %;
electrical engineering: 9.5 %; precision mechanics and optics: 3.1 %; office machines and data processing
equipment: 11.9 % (see BMBF, 1996).
(2) Total employment share of people with university education in West Germany (without public sector; 1995):
6.5 %; financial services: 7.8 %; legal and management services: 30.5 %; engineering, architectural, research
and testing services: 37.7 %; advertising: 20.4 % (see BMBF, 1996; 1997).
209
The job creation potential of the service sector in Europe
Appendix 3: Tables and figures
Agglomerated areas in Germany
Name of the agglomerated areas
The following employment office districts
(Arbeitsamtsbezirke) are included
1.1.1.1.3. Hamburg
Bremen
Hanover
Essen
Düsseldorf
Cologne
Frankfurt/Main
Rhine-Neckar
Stuttgart
Nuremberg
Munich
Hamburg, Bad Oldesloe, Elmshorn, Lüneburg
Bremen, Bremerhaven, Oldenburg, Verden
Hanover, Celle, Hameln, Nienburg
Essen, Bochum, Gelsenkirchen, Oberhausen
Düsseldorf, Wuppertal
Cologne, Bergisch Gladbach, Bonn, Brühl, Düren, Solingen
Frankfurt/Main, Darmstadt, Hanau, Offenbach, Wiesbaden
Heidelberg, Ludwigshafen, Mannheim, Neustadt
Stuttgart, Göppingen, Ludwigsburg. Waiblingen
Nuremberg, Bamberg, Weißenburg
Munich, Freising, Rosenheim
Table 1. Changes in employment in West Germany, 1977–98 (%) (1)
West Germany
Job loss
Additional jobs
Balance (total)
Proportion from women of
Job loss
Additional jobs
Balance (total)
Relation (2)
Job loss
Agriculture/mining
Knowledge-intensive industries
Other industries
Construction
Additional jobs
Services
Business services
Knowledge-intensive services
Finance and insurance
Other business services
Distribution services
Consumer services
Social services
Government services
Services
Domestic-trade-oriented services
Expert-oriented services
Agglomerated areas Non-agglomerated areas
– 8.5
19.8
11.3
– 11.5
19.3
7.8
– 6.3
20.5
14.2
31.3
64.3
91.7
2.32
100
9.0
15.8
62.6
12.6
29.9
59.9
108.5
1.68
100
0.1
20.2
66.1
13.6
33.2
67.8
83.8
3.36
100
21.9
9.5
57.2
11.4
100
36.4
21.7
5.7
14.7
14.4
5.9
42.3
1.0
100
74.1
25.9
100
45.5
27.1
6.2
18.4
9.8
6.4
38.0
0.3
100
69.2
30.8
100
29.0
17.4
5.3
11.6
18.2
5.5
45.9
1.4
100
78.2
21.8
(1) See Appendix 2.
(2) Relation between job loss/additional jobs.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
210
Determinants of business and personal services: Evidence from the German regions
Table 2. Changes of employment shares in West Germany, 1977–98 (%)
West Germany
Agriculture/mining
Industry
Knowledge-intensive industries
Other industries
Construction
Services
Services (total)
Business services
Knowledge-intensive services
Other services
Distribution services
Consumer services
Social services
Government services
Services (total)
Domestic-trade-oriented services
Expert-oriented services
Agglomerated
areas
Non-agglomerated
areas
1977
1998
1977
1998
1977
1998
100
100
100
100
100
100
3.5
42.5
16.0
26.5
8.0
45.9
100
16.8
12.0
4.8
40.5
8.5
20.0
14.2
100
79.9
20.1
2.5
32.6
13.3
19.3
6.2
58.6
100
22.7
15.1
7.6
32.9
7.5
26.6
10.3
100
76.4
23.6
2.9
39.2
17.8
21.4
7.2
50.7
100
19.5
14.0
5.5
41.2
8.3
18.3
12.7
100
76.7
23.3
2.1
28.3
13.7
14.6
5.3
64.2
100
26.9
17.9
9.0
33.0
7.9
23.2
9.0
100
72.7
27.3
4.1
45.7
14.2
31.5
8.6
41.6
100
13.7
10.0
3.7
41.5
8.7
21.8
14.3
100
83.2
16.8
2.8
36.2
12.7
23.5
6.9
54.1
100
18.7
12.4
6.3
32.7
7.6
29.7
11.3
100
80.0
20.0
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 1. Employment trends in West Germany, 1977–98 (%) (1)
(1) Job loss, additional jobs and balance in per cent of total employment in 1977. For agriculture/mining,
knowledge and other industries, construction and services: employment change between 1977 and 1988.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
211
The job creation potential of the service sector in Europe
Figure 2. Employment trends in West Germany, 1977–89 (%) (1)
(1) See Figure 1.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 3: Employment trends in West Germany, 1989–98 (%) (1)
(1) See Figure 1.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
212
Determinants of business and personal services: Evidence from the German regions
Figure 4: Employment trends in agglomerated areas, 1989–98, (%) (1)
(1) See Figure 1.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 5. Employment trends in non-agglomerated areas, 1989–98 (%) (1)
(1) See Figure 1.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
213
The job creation potential of the service sector in Europe
Table 3: Changes of employment in West Germany, 1977–98, Women (%)
West Germany
Job loss
Additional jobs
Balance (total)
Relation (1)
Job loss
Knowledge-intensive industries
Other industries
Additional jobs
Agriculture/mining
Construction
Services
Business services
Knowledge-intensive services
Finance and insurance
Other business services
Distribution services
Consumer services
Social services
Government services
Services
Domestic-trade-oriented services
Expert-oriented services
Agglomerated areas Non-agglomerated areas
– 6.1
33.3
27.2
– 7.4
29.7
22.3
– 5.1
37.6
32.5
4.55
100
23.3
36.7
100
0.6
1.4
98.0
27.3
17.3
5.4
9.4
12.9
4.6
47.2
6.0
100
78.0
22.0
3.47
100
38.4
61.6
100
0.6
0.8
98.6
35.0
22.5
6.1
12.5
9.1
5.3
44.5
6.1
100
72.2
27.8
5.82
100
5.6
94.4
100
0.6
1.9
97.5
21.8
14.7
4.9
7.1
15.9
4.1
49.2
9.0
100
79.5
20.5
(1) Relation between job loss/additional jobs.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
214
Determinants of business and personal services: Evidence from the German regions
Table 4: Changes of employment in West Germany 1977–98: Men (%)
West Germany
Job loss
Additional jobs
Balance (total)
Relation (1)
Job loss
Agriculture/mining
Knowledge-intensive industries
Other industries
Construction
Government services
Additional jobs
Services
Business services
Knowledge-intensive services
Finance and insurance
Other business services
Distribution services
Consumer services
Social services
Services
Domestic-trade-oriented services
Expert-oriented services
Agglomerated areas Non-agglomerated areas
– 10.1
11.6
1.5
– 13.1
12.0
– 1.1
– 7.6
11.2
3.6
1.15
100
13.2
10.3
47.6
19.7
9.2
0.99
100
1.0
10.8
62.6
19.4
6.2
1.37
100
30.0
9.6
26.9
20.1
13.4
100
48.5
26.0
5.5
22.5
15.4
7.5
28.6
100
61.3
38.7
100
57.5
31.7
5.7
25.8
10.1
7.5
24.9
100
59.0
41.0
100
39.6
20.4
5.3
19.2
20.7
7.5
32.2
100
63.6
36.4
(1) Relation between job loss/additional jobs.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
215
The job creation potential of the service sector in Europe
Figure 6: Employment trends in West Germany, 1977–98, Women (%)
(1) See Figure 1.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 7: Employment trends in West Germany, 1977–98, Men (%)
(1) See Figure 1.
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
216
Determinants of business and personal services: Evidence from the German regions
Figure 8: Employment trends in services in West Germany, 1977–89 (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 9: Employment trends in services in West Germany, 1989–98 (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
217
The job creation potential of the service sector in Europe
Table 5: Changes of employment shares in West Germany, 1977–98 Women (%)
West Germany
Agriculture/mining
Industry
Knowledge-intensive industries
Other industries
Construction
Services
Services (total)
Business services
Knowledge-intensive services
Finance and insurance
Other services
Distribution services
Consumer services
Social services
Government services
Services (total)
Domestic-trade-oriented services
Expert-oriented services
Non-agglomerated
areas
1977
1998
1977
1998
1977
1998
100
100
100
100
100
100
1.2
32.9
11.0
21.9
1.5
64.4
100
16.4
11.6
7.5
4.8
34.5
10.2
27.0
11.9
100
83.8
16.2
1.1
19.6
7.2
12.4
1.6
77.7
100
20.1
14.0
6.8
6.1
27.3
8.3
34.5
9.8
100
81.8
18.2
1.1
29.1
12.2
16.9
1.5
68.3
100
18.9
13.7
9.1
5.2
35.8
9.4
24.1
11.8
100
80.7
19.3
1.0
16.3
7.1
9.2
1.4
81.3
100
24.1
16.5
8.2
7.6
27.5
8.1
30.7
9.6
100
78.4
21.6
1.4
36.3
10.0
26.3
1.5
60.8
100
13.0
9.6
5.9
3.4
34.0
11.1
29.9
12.0
100
86.8
13.2
1.2
22.3
7.3
15.0
1.7
74.8
100
16.6
11.7
5.6
4.9
27.1
8.4
37.8
10.1
100
84.8
15.2
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
218
Agglomerated
areas
Determinants of business and personal services: Evidence from the German regions
Table 6: Changes of employment shares in West Germany, 1977–98: Men (%)
West Germany
Agriculture/mining
Industry
Knowledge-intensive industries
Other industries
Construction
Services
Services (total)
Business services
Knowledge-intensive services
Finance and insurance
Other services
Distribution services
Consumer services
Social services
Government services
Services (total)
Domestic-trade-oriented services
Expert-oriented services
Agglomerated
areas
Non-agglomerated
areas
1977
1998
1977
1998
1977
1998
100
100
100
100
100
100
4.9
49.0
18.9
30.1
11.8
34.3
100
17.6
12.8
8.4
4.8
44.4
6.4
11.4
20.2
100
70.3
29.7
3.5
42.4
17.6
24.8
9.7
44.4
100
26.1
16.6
7.8
9.5
40.2
6.8
16.1
10.8
100
69.4
30.6
3.2
46.9
21.9
25.0
10.8
39.1
100
20.6
15.0
9.9
5.6
48.7
6.9
10.6
13.2
100
66.8
33.2
3.1
37.3
20.5
16.8
8.3
51.3
100
30.3
19.5
9.0
10.8
39.6
7.1
14.4
8.6
100
65.9
34.1
6.4
50.7
16.4
34.3
12.7
30.2
100
14.5
10.5
6.9
4.0
46.6
5.9
12.3
20.7
100
74.0
26.0
3.9
46.5
15.2
31.3
10.8
38.8
100
21.7
13.5
6.6
8.2
40.8
6.5
18.0
13.0
100
73.1
26.9
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
219
The job creation potential of the service sector in Europe
Table 7: Employment from women as a proportion of total employment (%)
West Germany
Total
Agriculture/mining
Industry
Knowledge-intensive industries
Other industries
Construction
Services
Business services
Knowledge-intensive services
Finance and insurance
Other services
Distribution services
Consumer services
Social services
Government services
Domestic-trade-oriented services
Expert-oriented services
Non-agglomerated
areas
1977
1998
1977
1998
1977
1998
37.3
13.0
28.6
25.8
30.2
7.0
52.8
50.3
50.3
49.9
50.5
45.1
64.2
72.7
43.9
57.2
38.0
42.8
19.5
25.7
23.5
27.2
10.9
56.7
50.3
52.5
53.2
46.0
47.1
61.4
73.7
54.4
60.7
43.9
38.0
17.2
27.7
25.6
29.4
7.8
51.9
49.9
49.6
50.0
50.6
44.3
59.5
71.1
48.8
56.6
38.6
43.1
20.4
24.9
20.8
29.3
11.4
54.5
48.8
50.4
52.2
45.7
45.4
57.7
71.8
57.3
58.7
43.2
36.7
11.2
29.2
26.0
30.7
6.5
53.7
51.0
51.3
49.9
50.3
45.9
68.6
73.9
40.0
57.7
37.1
42.6
18.9
26.3
26.2
26.3
10.8
58.9
52.4
55.4
54.6
46.2
48.7
65.0
75.1
52.3
62.4
44.8
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
220
Agglomerated
areas
Determinants of business and personal services: Evidence from the German regions
Figure 10: Employment trends in services in West Germany, 1989–98: Women (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 11: Employment trends in services in West Germany, 1989–98: Men (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
221
The job creation potential of the service sector in Europe
Figure 12: Industry and total employment: growth rate, 1977–89 (%)
r = 0.90
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 13: Industry and total employment: growth rate, 1989–98 (%)
r = 0.69
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
222
Determinants of business and personal services: Evidence from the German regions
Figure 14: Industry and service employment: growth rate, 1977–89 (%)
r = 0.69
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 15: Industry and service employment: growth rate, 1989–98 (%)
r = – 0.25 (agglomerated areas alone: — 0.49)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
223
The job creation potential of the service sector in Europe
Figure 16: Industry and business services: growth rates, 1977–89 (%)
r = 0.42
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 17: Industry and business services: growth rates, 1989–98 (%)
r = – 0.23 (agglomerated areas alone: – 0.42)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
224
Determinants of business and personal services: Evidence from the German regions
Figure 18: Knowledge-intensive industries/business services and total employment: growth rate,
1977–89 (%)
r = 0.62
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 19: Knowledge-intensive industries/business services and total employment: growth rate,
1989–98 (%)
r = 0.70
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
225
The job creation potential of the service sector in Europe
Table 8: Shift-share components of service employment change in West Germany, 1977–98 (%)
Locational component
Structural component
Total employment
change
Nuremberg
Munich
Rhein/Neckar
Stuttgart
Cologne
Frankfurt
Düsseldorf
Hanover
Bremen
Essen
14.2
5.9
1.9
0.7
0.0
– 1.5
– 9.0
– 11.8
– 12.6
– 18.4
39.2
47.5
43.6
43.8
45.5
45.7
42.7
44.2
37.0
43.5
53.4
53.4
45.5
44.5
45.5
44.2
33.7
32.4
24.4
25.1
Hamburg
Agglomerated areas (total)
Non-agglomerated areas
West Germany
– 19.2
– 8.2
7.5
0.0
41.3
46.9
41.2
43.6
22.1
38.7
48.7
43.6
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Table 9: Shift-share components of service employment change in West Germany, 1977–89 (%)
Locational component
Stuttgart
Munich
Nuremberg
Frankfurt
Rhein/Neckar
Düsseldorf
Cologne
Bremen
Hanover
Hamburg
Essen
Agglomerated areas (total)
Non-agglomerated areas
West Germany
8.1
7.4
5.5
1.2
0.2
– 2.2
– 4.0
– 5.2
– 6.5
– 10.3
– 11.9
– 1.9
3.1
0.0
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
226
Structural component
Total employment
change
20.7
24.6
22.8
23.4
23.1
22.0
23.8
19.9
23.3
21.3
22.7
22.7
22.2
23.0
28.8
32.0
28.3
24.6
23.3
19.8
19.8
14.7
16.8
11.0
10.8
20.8
25.3
23.0
Determinants of business and personal services: Evidence from the German regions
Table 10: Shift-Share components of service employment change in West Germany, 1989–98 (%)
R 89/98
Cologne
Nuremberg
Rhein/Neckar
Frankfurt
Munich
Hanover
Essen
Stuttgart
Düsseldorf
Hamburg
Bremen
Agglomerated areas (total)
Non-agglomerated areas
1
2
3
4
5
6
7
8
9
10
11
R 77/89
Locational
component
Structural
component
Total
employment
change
7
3
5
4
2
9
11
1
6
10
8
4.1
3.2
1.1
– 1.2
– 1.7
– 3.2
– 4.4
– 4.5
– 5.1
– 6.4
– 6.8
– 2.1
3.2
17.3
16.2
16.9
16.9
17.9
16.6
17.3
17.1
16.7
16.3
15.3
16.9
15.5
21.4
19.0
18.0
15.7
16.2
13.4
12.9
12.6
11.6
9.9
8.5
14.8
18.7
0.0
16.7
16.7
West Germany
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 20: Correlation between the structural components of service employment change and the
total service employment change, 1977–89 (%) r = 0.38
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
227
The job creation potential of the service sector in Europe
Figure 21: Correlation between the structural components of service employment change and the
total service employment change, 1989–98 (%) r = 0.51
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 22: Correlation between the locational components of service employment change and the
total service employment change, 1977–89 (%) r = 0.98
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
228
Determinants of business and personal services: Evidence from the German regions
Figure 23: Correlation between the locational components of service employment change and the
total service employment change, 1989–98 (%) r = 0.97
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
229
The job creation potential of the service sector in Europe
Table 11: Relations of employment rates, 1998, between West Germany, Düsseldorf, Munich and
Stuttgart (employees in percent of inhabitants)
West Germany
Total
Industries
Knowledge-intensive industries
Other industries
Services
Business services
Knowledge-intensive services
Other business services
Distribution services
Consumer services
Social services
Government services
Knowledge-intensive industry/services
Household related services
Domestic-trade-oriented services
Expert-oriented services
Düsseldorf
Munich
Stuttgart
36.2
43.5
40.3
39.5
100
10.2
100
5.0
100
5.3
100
23.2
100
6.2
100
4.2
100
2.0
100
7.7
100
1.8
100
5.4
100
2.1
100
120
12.0
117
4.8
97
7.2
136
29.2
126
8.4
134
5.6
136
2.8
131
10.9
142
1.9
107
5.3
98
2.8
131
111
9.6
94
5.5
111
4.1
78
27.8
120
8.1
130
6.0
144
2.1
103
8.2
107
3.0
168
6.4
119
2.8
131
109
16.1
157
9.6
193
6.5
123
20.8
90
6.0
95
4.3
104
1.7
79
6.5
85
1.6
91
4.5
84
2.1
100
9.1
100
7.3
100
16.9
100
6.3
100
10.4
115
11.2
153
20.8
123
8.4
133
11.5
126
10.2
140
20.4
121
7.4
117
13.9
152
7.4
101
14.8
88
6.0
95
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
230
Determinants of business and personal services: Evidence from the German regions
Table 12: Relations of employment rates, 1998, between West Germany, Bremen and Essen
(employees in percent of inhabitants)
West Germany
Bremen
Essen
36.2
27.5
26.7
100
10.2
100
5.0
100
5.3
100
23.2
100
6.2
100
4.2
100
2.0
100
7.7
100
1.8
100
5.4
100
2.1
100
9.1
100
7.3
100
16.9
100
6.3
100
63
8.4
82
3.2
64
5.2
99
20.3
87
4.0
63
2.4
58
1.6
74
7.7
100
1.1
61
5.5
102
2.0
95
5.6
61
7.6
104
15.5
92
4.8
75
59
7.3
72
2.6
52
4.8
90
19.4
83
4.3
69
2.6
62
1.7
79
6.4
83
1.2
65
5.8
107
1.6
76
5.1
56
7.3
100
15.3
91
4.0
64
Total
Industries
Knowledge-intensive industries
Other industries
Services
Business services
Knowledge-intensive services
Other business services
Distribution services
Consumer services
Social services
Government services
Knowledge-intensive industry/services
Household related services
Domestic-trade-oriented services
Expert-oriented services
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
231
The job creation potential of the service sector in Europe
Table 13: Employment shares in knowledge-intensive industries, 1998 (%)
Employment total
Industries
Knowledge-intensive industries
Knowledge-intensive industries as
percent of industrial employment
Chemicals
Machinery
Motor vehicles
Aircraft
Electronic
Computing hardware
Instruments/optics
Employment total
Industries
Knowledge-intensive industries
Knowledge-intensive industries as
percent of industrial employment
Chemicals
Machinery
Motor vehicles
Aircraft
Electronic
Computing hardware
Instruments/optics
West Germany
Düsseldorf
Munich
Stuttgart
100
32.6
13.1
100
27.5
11.1
100
23.9
13.8
100
40.7
24.2
40.2
6.7
7.8
9.1
0.8
12.3
0.7
2.8
40.3
13.6
7.4
4.9
0.2
12.0
0.2
1.7
57.8
6.0
4.4
13.2
5.8
21.5
3.0
3.9
59.6
2.3
11.4
22.8
0.1
17.7
0.1
3.4
West Germany
Bremen
Essen
100
32.6
13.1
100
26.5
10.1
100
24.1
8.5
40.2
6.7
7.8
9.1
0.8
12.3
0.7
38.0
2.6
5.9
12.9
4.3
9.7
0.2
35.2
4.8
5.2
11.9
0.0
11.9
0.2
2.8
1.1
2.0
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
232
Determinants of business and personal services: Evidence from the German regions
Table 14: Employment shares in knowledge-intensive business services, 1998 ( %)
West Germany
Düsseldorf
Munich
Stuttgart
Employment total
Services
Business services
Knowledge-intensive
business services
Knowledge-intensive
business services as percent of
business service employment
Finance and insurance
Legal and management services
Engineering, architectural services
100
58.6
13.1
100
67.1
19.2
100
69.3
20.3
100
52.6
15.1
8.1
12.9
14.9
10.9
66.4
32.4
18.4
12.8
67.2
31.1
21.7
7.3
73.4
34.2
15.4
20.5
72.3
34.4
19.4
14.6
Public relations, advertising services
2.8
7.1
3.3
3.9
West Germany
Bremen
Essen
Employment total
Services
Business services
Knowledge-intensive
business services
Knowledge-intensive
business services as percent of
business service employment
Finance and insurance
Legal and management services
Engineering, architectural services
100
58.6
13.1
100
64.2
12.5
100
63.8
14.3
8.1
7.7
8.4
66.4
32.4
18.4
12.8
60.8
29.1
18.2
11.1
58.9
19.9
21.3
14.7
Public relations, advertising services
2.8
2.4
3.0
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
233
The job creation potential of the service sector in Europe
Figure 24: Changes of employment shares in Düsseldorf, 1977–98 (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 25: Changes of employment shares in Munich, 1977–98 (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
234
Determinants of business and personal services: Evidence from the German regions
Figure 26: Changes of employment shares in Bremen, 1977–98 (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 27: Changes of employment shares in Essen, 1977–98 (%)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
235
The job creation potential of the service sector in Europe
Figure 28: Correlation between employment shares in business services and the total employment
rate, 1998
r = 0.65
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 29: Correlation between employment shares in knowledge-intensive business services and the
total employment rate, 1998
r = 0.70
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
236
Determinants of business and personal services: Evidence from the German regions
Figure 30: Correlation between employment shares in knowledge-intensive industries/business
services and the total employment rate, 1998
r = 0.56
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
Figure 31: Household related services and total employment: growth rates, 1977–98 (%) r = 0.77
(without Essen: 0.63)
Source: Regionaldatenbank Arbeitsmarkt am WZB/AB.
237
Chapter 3
Analysing business services employment —
some theoretical and methodological remarks
Bernard Gazier and Nadine Thevenot
3.1. Introduction
With the development of the business service sector, many hopes arose regarding job creation
in European countries. The main reasons are linked to the secular growth of service
employment, in particular to the growth of employment in market services compared with the
decline of agricultural and industrial employment. In the last decades, it is the business service
sector within the broader service sector that many economists and specialists have found
attractive because of its rapid growth. One illustration of this is in a recent report provided by
the main French economic institute on employment in the market services sector: business
services comprise 60 % of all jobs created in the market service sector between 1962 and 1996
in France (INSEE, 1999). Several explanations of this evolution are usually suggested: (1)
some services are new services that respond to new needs for user firms (like computer
services); (2) more usual service activities performed in the past by user firms’ employees are
now externalised (cleaning, security, accounting and legal functions, etc.); (3) firms develop
new ways of organising work and activities to adapt to new conditions of competitiveness like
search of flexibility.
The purpose of this chapter is to explore, in a theoretical and methodological way, some
important dimensions and determinants of such a growth. A fundamental problem raised is
that numerous processes interact in the development of services because of their relational
character. Generally, services involve a more or less close relationship between customer and
producer, services may be organised in various ways, with various consequences on the
nature, status and scope of the work relationship. The question of the quantity of jobs created
interferes with the question of the very nature of these jobs: salaried or self-employed, stable
or precarious?
239
The job creation potential of the service sector in Europe
In the precise case of business services, these questions are magnified. As stated above, firms
are service users but they can choose to produce or buy them. So the competitive pressures
probably are stronger than in other service sectors (even if some tough competition can be
found in private personal services) because the option of performing the task in-house is
systematically considered by the firms and they look for competitive advantages in each
dimension of their activity.
These questions are complicated by the externalisation process evoked above. When one tries
to assess the job creation potential of the business service sector, one faces the problem of job
transfers.
Varied approaches can be found in the literature, stressing a wide span of factors. Some recent
contributions may be of interest here, either stressing the potential for numerous, high-quality
new jobs in a positive way, or insisting on the risk of fewer, unstable and low-quality jobs.
A systematic confrontation being out of our reach, we will limit ourselves to some clearly
identified recent approaches. They will be diverse, belonging to different disciplines:
economic approaches from various fields (labour, industrial and organisational economics), a
management approach and labour law approach. Some approaches are a combination of
disciplines.
Our design is to introduce three sources of complexity step-by-step, and to discuss their
employment consequences in the case of business services. The first is the need for trust and
loyalty. The second is a diversified process of risk management underlying various work
arrangements, and the third is the heterogeneity of the very nature of services provided.
We begin by analysing a purely managerial point of view, which is optimistic for services as
a whole (Section 3.2). We will show that the main argument, centred on the need to build a
long-lasting relationship of trust and loyalty between providers and clients, strongly depends
on a double segmentation process affecting customers and workers. We briefly apply such
reasoning to business services.
In the second step, we keep the concern of trust and loyalty, and introduce a recent labour law
and economic risk treatment approach, developed for any kind of work relationship but with
strong applications to business services. The main result is the identification of increasingly
blurred frontiers between salary and self-employed work, and considerable room for
manoeuvring as regards the organisational and employment choices of firms (Section 3.3).
Then we focus exclusively on the case of business services, and explore an economic
approach in more detail, combining the labour economics, industrial economics and
organisation economics points of view. Some dimensions of service heterogeneity (tangibility,
specificity, degree of user’s participation) are considered since they imply different risks and
stakes and help us understand their various externalisation potential (Section 3.4).
240
Analysing business services employment — some theoretical and methodological remarks
Last, as an application of the previous developments, we deal briefly with some persistent
problems affecting the measurement of externalised business service employment, and we
present some hypotheses on the differentiated development of such services (Section 3.5).
3.2. An optimistic management approach: The ‘service profit chain’
The focus of this section is on assessing the scope and validity of a recent management book
centred on the idea of a ‘service profit chain’. (1)
The book deserves special attention for three reasons. First it is written by three renowned
specialists of management. Second, it is the converging point of previous research, which was
done somehow with other collaborators. The book is synthetic in its aim — it seeks to
generalise a set of relationships. Third, the rather optimistic message is based on an integrated
set of relatively new arguments. The main thesis is that there is a ‘mirror effect’ between
loyalty and satisfaction of employees. When this occurs it is possible to build a stable, longlasting relationship which generates profitable business in services. The ‘service profit chain’
is the link between long-standing customers and employees, and growth and profit for the
service firm.
The message seems to be common sense and is certainly good news: well-paid and valued
employees lead to high-quality service and profits. Yet the book goes further, notably because
it insists more on employee attitudes and values than on skills or diplomas. This kind of
diagnosis is in line with the idea that a new ‘service economy’ might develop, offering good
jobs with initiative, responsibility and high rewards to the bulk of employees, not only to a
well-trained elite.
Three main strands of arguments are combined in the book, each deserving attention.
1. The authors develop an operational analysis of the relatively complex processes existing in
a service relationship, starting from the consumer’s point of view and combining it with the
employee’s point of view. Two sources of complexity interact. First, consumer diversity:
different people expect different quality. Second, service heterogeneity: in fact a service
relationship can combine very different elementary operations; some of them directly done
by an employee, others by a machine, and others by some mix of machine and man.
According to the situations and preferences, the combination could evolve. The main
operational consequence is the need of segmenting customers. A central idea in the book,
is that the customer is not always right: some of them could ask for a service or a service
quality not belonging to the target of a given service business, and they need to be
reoriented towards other firms.
(1) Heskett, Sasser, and Schlesinger, 1997, The service profit chain. How leading companies link profit and growth
to loyalty, satisfaction and value, The Free Press, 301 pp.
241
The job creation potential of the service sector in Europe
2. Then, with regard to the employees, there is a need to foster initiative, especially for
frontline workers, involved in the ‘service encounter’. They must be able to select a service
subset by themselves and adapt quickly to the demand. The second operational
consequence is giving them a ‘capability’, namely, some latitude in performing their work
(within specified limits), and a number of disposable means and background technical
support. A third consequence is the need for selecting employees according to the
corresponding profile: ‘nice’ people are recruited; those who are open-minded and flexible.
The customer segmentation is mirrored by an employee segmentation. The technical skills
and the training policies, though important, are presented here as being secondary.
3. If we put together the two preceding lines of argument, we find the organisational basis for
a stable, long-lasting relationship between a firm and a service customer. The book insists
on the ‘lifetime value of a customer’, namely, the revenue flow stemming from a satisfied
customer maintaining an active participation through the years. As most of the service
relationships are personal, frequent, and involve some risks, it is important and profitable
to build mutual trust, then a satisfied customer consumes an evolving set of services (for
example, in a banking activity), and generates other clients with limited additional costs.
However, some unaddressed questions arise when one tries to assess the general lessons
stemming from the book in more detail.
What is first lacking is a specific application to business services. The authors do not focus on
such services even if they do not exclude them. For example, they discuss the case of pest
exterminators, a service with important clients in businesses like restaurants or hotels. It seems
reasonable to state that the business service sector is fully covered by the arguments presented
in the book. Indeed, business services exacerbate the stakes and risks of consumer
dissatisfaction. As was noted above, another alternative always exists — in-house production.
The competitive pressure is probably tougher, because an ill-performed service may have
important negative consequences for the efficiency of the client, and because the client is
trained at looking for available alternatives and minimising costs. The consequence yields
more competitive pressures, on more dimensions. It may lead to frequent checks, and also to
systematic networking: either the systematic research of a price advantage, or a closely
matched network of firms sharing the same values and technological level.
The authors illustrate four extreme behaviours of clients (p. 85). The ‘apostles’ are strongly
satisfied and loyal, the ‘terrorists’ are neither satisfied nor loyal, the ‘hostages’ are not satisfied
but loyal (because they are, for example, hostages of a monopoly), and last the ‘mercenaries’
are satisfied but not loyal. In the case of business services, there is an overall tendency leading
to more ‘mercenary’ behaviour. ‘Hostages’ and ‘terrorists’ may exist, of course, but the need
for some quality and satisfaction as well as the private operating context (2) may limit their
importance. ‘Apostles’ may become rare, except in the case of high-quality level networks.
(2) Things could be very different in a developing country context, with some business services in a monopoly
situation (for example, the software and computer maintenance services).
242
Analysing business services employment — some theoretical and methodological remarks
The book’s argument leads to similar conclusions, which paint a more unstable picture. Even
if trust is a key component, it leads to a contrasted balance with loyalty: less loyalty in the
sector (and higher price sensitivity of clients) and more intense loyalty (among peers) when it
is obtained.
We now turn to methodological remarks. It must be first observed that with this book we have
the classical advantages and drawbacks of most management analyses. They are based on
concrete and operational experiences, and they focus on integrated relationships (such as the
interaction between employees and customers). However, they give only limited grounds for
easily generalising. The method followed in the book is threefold. Firstly, it presents some
comparative evidence for ten big service firms using methods that could be connected with
the management proposals of the ‘service profit chain’, these firms being much more
profitable than a selected sample of firms bringing ‘reasonable matching comparisons’,
(p.15). Secondly, it documents some correlations between variables belonging to the
conceptual ‘service profit chain’. Yet these correlations, although presented in a systematic
table (p. 31), are partial correlations extracted from different cases. Lastly, the book presents
a wide number of case studies.
There appear to be two main limitations found in the analysis, from an employment-centred
point of view.
The first concerns the size and frequency of the customer and or employee segmentation. The
book gives numerous evidence of self-reinforcing links: some of them within the ‘service
profit chain’, each variable reinforcing others, and some of them within the ‘employee and or
customer mirror’, good quality ‘winning’ employee attracting good quality (as regards the
target of the service business) and ‘winning’ customers. It must be noticed, however, that the
book (logically) limits the segmentation process and the segmentation needs of frontline
employees. However, the study does not address the question of the effective growth of such
a category. What about the backline employees? What do they need or what do they tolerate?
A second question is how to go further as regards the organisational aspects of the service
relationship. The need to build trust in difficult-to-control, frequent and risky interactions,
evokes the williamsonian theory. However, here, such a situation is used only to introduce the
need for some contractual stability between stable employees and customers; that is not the
‘hierarchy’ concept. The book insists on the need for involving the frontline employees in the
business gains (participation in capital and management) as well as in profit-related gains.
A wider management approach is in order, with other ways of building trust without
necessarily engaging the firm in long-term employment relations.
243
The job creation potential of the service sector in Europe
3.3. Combining a labour law and economic approach: Risk management,
risk spillover and the service employment relationship
The ground upon which the rather optimistic conclusions of the ‘service profit chain’ were
built, emphasises the needed continuity and trust of the service relationship. In a sense, the
authors went back to the classical ‘internal labour markets’ developed by big firms during the
1960s in industrialised countries, with the aim of stabilising important groups of employees
and offering them good employment perspectives. However, their proposals differ in two
ways. They focus on ‘frontline’ employees, who are directly in contact with clients, and they
insist on the need to establish individual autonomy and performance-related compensation.
An important aspect was ignored in these developments: if the product cannot be dissociated
from the actual performance of the worker (namely, one of the main definitions of services),
then the burden of establishing the continuity and trust, of segmenting the customers and
selecting some of them for a durable interaction, of adapting to their evolving needs, usually
shared by the firm and the worker, may in some cases, be transferred to other organisations or
to the worker alone. This results in the worker being directly responsible for the long-lasting
relationship with the customer without involving the main firm in a stable employment
relationship. Some small subcontractors may then be as dependent or insecure as temporary
salaried workers, and at the same time, some employed workers may be treated as if they were
self-employed, responsible for their activities.
An interesting analysis, along these lines, was recently presented in collective research which
centred on business service activity under the direction of M. L. Morin. This work joined the
economic-management and labour law points of view. The aim was to help clarify the
increasingly blurred borderlines between independent and salaried work in France (3). The
method was based upon a conceptual (economic and labour law) elaboration of work relations
and in-depth case studies regarding recent developments and practices; some were explicit,
others more or less hidden or implicit.
They first presented: (a) a general assessment of the various risks which occur in a work
situation (either self-employed or employed). Then they showed, in the domain of business
services: (b) that numerous management choices must be analysed as intermediate choices
between self-employment and pure employment, depending on organisational choices and
risk sharing choices.
Turning back to the work relationship, four risk categories were distinguished:
— economic risks
— employment risks
— work security risks
— social risks
(3) Morin (ed,) 1999, Prestation de travail et activité de service, La Documentation Française.
244
Analysing business services employment — some theoretical and methodological remarks
The economic risks are business risks: financial risk of failure, of profit and losses, of
shareholder pressure, risk of irregular demand, and of production errors. These risks are
classically borne by the owner of the firm.
The employment risks are the risks stemming from creating and filling a job: they include the
job’s responsibilities, the job’s duration, and the employability of the worker (ability to find
another job). Such risks are traditionally shared by the employer and the community. In some
cases they may involve an important share of individual responsibility and management (for
example, reputational risks for the employed recruited on a short-term project basis).
The work security risks are the risks of work accidents which are individually and collectively
borne by the employer. Finally, the social risks arise in the situations where individual or
family needs of the worker cannot be covered by current labour income: sickness, old age,
childcare and unemployment. These risks are traditionally borne by the community (social
security arrangements).
Thus the starting situation puts the salaried worker and the self-employed worker in
opposition to each other. The economic risks and the security risks are mainly borne by the
employer, either solely or collectively, while the two remaining risks, employment risks and
social risks, are mainly shouldered by the workers, primarily in a collective manner. For the
self-employed, everything is the responsibility of the worker, either individually or
collectively.
However this elementary sharing has recently undergone changes, especially in the case of
business services. Two main hypotheses are offered about risk management today.
(1) There is a clear tendency to transfer more responsibility to the salaried worker: it amounts
to a risk spillover. For example, rewarding individual performance when it depends on
unpredictable market conditions, is a risk transfer (economic risks) from the firm to the
worker. Another example could be policies aimed at strengthening the individual
initiatives for ‘employability’. Here, it is the employment risks that are, in part, transferred
to the worker.
(2) This tendency does not lead to unilateral changes in the employment relationship, because
it must be combined with a variety of organisational choices. Risk management leads to
very different work arrangements, even for the same task. It can be performed within
different organisational frameworks, with different sharing of risks, duties and
compensations.
The main contribution of the research appears here: a systematic table allowing the situation
of hybrid arrangements between the classical salaried work contract and self-employed status.
245
The job creation potential of the service sector in Europe
Table 3.1. Risk management and work relationship
Risk of the service
provision Yi
Yαi + βc*
Yc
Work organisation
(individual)
(mixed)
(collective)
1
Self-employment
2
4
7
5
8
3
6
9
Pure salaried work
Xi
(individual)
Xαi +βc*
(mixed)
Xc
(collective)
* The α and β coefficients (α + β = 1) indicate that the proportions of the individual (i) and
collective (c) dimensions are mixed and may vary according to the different cases
introduced. (Cf. text.)
We first give some examples of the contents of the table, all taken from the business service
sector. Then we extract some more general statements about the span of service work
relationships, and the consequences as regards the kind of employment provided.
The best is to start from cell 9: pure salary work situation, with the firm acting as the collective
responsible. The odds of producing, pricing and trading are born by the firm, which also
benefits from the possible profits. If we look at cell 6, we find any individual performance
related to compensation for ‘normal’ employees; if we look at cell 8, we find a more
individualised work organisation. It could be separate software services within firms, outside
project missions for subcontracted workers, with clear work specification, or outside complex
and integrated cooperative projects for salaried workers. It could also be accounting inside the
firm, with work autonomy, as well as the separate training unit inside the firm. In each case,
the worker belongs to the firm as a salaried worker, but he or she performs more or less
independently of the core group.
If we cumulate the characteristics of cells 6 and 8, we find hybrid working arrangements:
separate tasks with individual performance-related compensation. Such a case, not the most
frequent, is observed in brand merchandising. The worker, hired by a specific trademark,
works inside a selling unit (department store) and represents his or her brand with personal
autonomy. The situation is analogous when an accounting expert, belonging to an accounting
firm, operates within a given client’s firm for a personal mission and manages the contract.
The work situation of cell 3 is quite frequent. Individual workers are integrated into a
collective work group, but are fully responsible for their service provision. It is the case of
some merchandising arrangements. The worker can be directed by the retail trade firm, but
hired and paid by the supplier of some product to be sold, on a purely individual basis (hours
worked). In the French contracts, it corresponds to ‘external aid paid by the supplier’. Another
246
Analysing business services employment — some theoretical and methodological remarks
similar case is any individual outside in a software project with permanent control of the
client.
At the opposite, cell 7 is quite rare and was not observed in the research. It corresponds to
workers with a salary contract but performing, for part of their activity, some managing task.
It happens in some small firms, where workers play two different roles.
Now we start from the self-employed worker (cell 1) and we introduce some complexity with
cell 2: more collective work organisation. It corresponds to repeated short-term tasks given by
training firms on the basis of a continuing flow of specific projects; either short-term salary
contracts, or short-term independent missions.
Lastly, cell 4 corresponds to some collective management of individual work: either by the
firm, for example, when a long-term contract is concluded for an independent software
assistant, or when an independent worker benefits from a public social protection scheme
allowing him/her to switch from one task to another with some income and insurance
guarantees, such is the case in France with the status of ‘intermittents du spectacle’.
This perspective brings an interesting broadening of the causality network surrounding the
choices made by firms as regards business services.
The main prospective orientation, left implicit by the authors, is that the cells on the diagonal
are increasingly active and correspond to various behaviours from the firms. The firms
transfer risks from themselves to outside subcontractors and/or to individualised workers,
either self-employed or benefiting as salaried, from increased autonomy. So the authors find
an interesting principle which helps us understand why the ‘service profit chain’ may lead to
out-of-the-firm and/or unstable work arrangements.
However, this approach gives only illustrations taken from different business service
activities, and does not attempt to explain why specific work arrangements may differ or
coexist in a given region and for a given service. The labour law point of view, combined with
a general risk-management argument, is enough to open a wide variety of arrangements, and
raise fundamental questions about the need of controlling and regulating the growing
dispersion stemming from the risk spillovers (4).
In order to discuss the consequences on job creation and job quality, it would be necessary to
consider in more detail the meaning of service heterogeneity from an organisational point of
view, and the issue of externalisation.
(4) A different but converging analysis is proposed in Benhamou and Gazier (cf. this research), about the
determinants of cultural employment. The aim of Morin et al. is to show that for a given service, different work
arrangements may coexist in a given country and a given place. However, the emphasis may be put in
contrasting national ‘employment regimes’ fostering different dominant work arrangements in each country.
Self-employment in British cultural employment is connected to weak labour market regulation, while the
importance of short-term contracts and part-time in the case of France is connected to the French specificity of
labour market regulation.
247
The job creation potential of the service sector in Europe
Thus, this study opens a highly general field, and needs complementary elements to give a
more detailed account of the various work relationships connected to service activity.
3.4. An economic approach: The determining factors of business service
externalisation
On a theoretical level, the theories of the firm develop two main lines of argument, reflecting
the traditional opposition in economic analysis: the treatment of economic phenomena in an
exchange or production point of view. In an exchange economy, we find the contractual
approaches, among them transaction cost theory to be the main ‘make or buy’ problem
treatment (Williamson, 1985). In a production economy, there is a set of contributions
stemming from the works of Penrose (1959) or Richardson (1972) that place resources and
competencies at the centre of the analysis. These recent works enter into the more general
evolutionary perspective by insisting on the condition of resource creation and accumulation.
These analyses do not specifically take into consideration services but suggest propositions
that are likely to be applied both to goods and services.
Though the arguments proposed by the theories of the firm are interesting for the study of the
organisation of service activities, it is necessary to consider the features of a given service to
understand the factors that explain why a firm performs the service in-house or uses external
service providers. The combination of industrial economics and service approaches allows us
to evaluate the features of service activities which are likely to affect their mode of
organisation (Section 3.4.1). In fact, these features concern the ‘object’ of the transaction and
the production and may change the main results of transaction and production analyses
(Section 3.4.2). Finally, we emphasise the different organisational arrangements by which
service activities can be provided. The classification underlines the respective roles of the
provider and the user in the conception and the production of a given business service. In this
perspective, we shall show that the alternative for the user firm is not that of ‘make or buy’ but
rather involves complex forms, which need both participation and competence in coproduction of service units internal to the user firm (Section 3.4.3).
3.4.1. The heterogeneity of business services: What are the origins?
We do not intend here to discuss the theoretical concept of services but rather aim at specifying
the characteristics of business services that explain their heterogeneity. From this point of
view, we consider business service as the result of a transformation process from the definition
of service as a change (Hill, 1977, Gadrey, 1992). It is the effect of this change that is expected
by firms asking for business services.
Characterising business service heterogeneity consists in differentiating the major dimensions
of service in combination with the components of service activity (Box 3.1).
248
Analysing business services employment — some theoretical and methodological remarks
Box 3.1. The components of business service activities
— the ‘expected useful effect’ of services for the user firm;
— the production process of services and the ‘service relationship’;
— the service as result;
— the realised effect of service.
This combination leads to the following characterisation (5):
— The expected useful effect refers to the ‘reach’ of services for the user firm, that is to the
operational or exploring character of service in the organisation of production of the
firm (6).
— The production process and the service relationship (7): The intensity of interactions
between user and provider is a determining factor of the definition and production of
service. The participation of the user firm can take place at several levels during the
provision of the service: the conception and definition of service, the production and
achievement of the service effects. The importance and the level of participation depend
on the specificity degree of the user firm needs in the sense that when the user firm has
particular needs, it has to define and produce with the provider a service which possesses
a high level of novelty.
— The service as a result is defined by a change in the condition of a ‘thing’ (Hill, 1977,
Gadrey, 1992) which is the ‘object’ of the transaction. The more this thing is intangible,
the more difficult the evaluation of the service (Barcet, Bonamy, 1994, Gadrey, Delaunay,
1987). It is interesting to differentiate services with the degree of tangibility of their
support.
— The realised effect of service has various temporalities for the user firm, that are the times
of using the service effect after the service has been produced and consumed. Moreover,
it is possible that the achievement timetable of service results does not fit in with the
timetable of service effects. This possible time lag enhances uncertainty for the user firm.
These dimensions lead us to consider several service categories that function differently in
production and transaction.
3.4.2. Business service activities and the ‘make or buy’ dilemma
In general, theories of the firm consider goods and services indifferently and do not take into
consideration the features of the ‘object’ of the production and transaction. Nevertheless, it
(5) For more details, cf. Thèvenot (1999).
(6) In the case of an ‘operational reach’, the firm is aiming to exploit an existing potential. An ‘exploring reach’
refers to services that are realised in association with the search of new productive opportunities.
(7) The ‘service relationship’ developed by French economists refers to the social and operational interactions
between user and producer in the production and organisation of both goods and services. Cf. De Bandt, Gadrey
dir. (1994), Reboud dir. (1997).
249
The job creation potential of the service sector in Europe
seems to us that these features have an impact on the main propositions of the two prevalent
contemporary approaches that deal with the firm’s boundary problems: the contractual and
competency perspectives (for a presentation of these analyses, see Box 3.2 in Appendix 1).
3.4.2.1. What are the transaction costs associated with service transactions?
Our intention is to show that the provision of service confronts producer and customer with
some particular contractual hazards linked to the features of this activity. The major
uncertainties affecting the transaction concern the evaluation of both service and provider
performances (i) and the appropriation of the result of the provision by the producer (ii). In
other respects, the fact that the user firm could have specific needs entails a decrease of
uncertainty due to the participation of the user firm in the provision of service (iii).
(i)
The evaluation of service and provider performances: Like many inter-firm
relationships, the provision of service is about a ‘product’ that does not exist before the
transaction. Thus, the contract that joins the firms is considered to be an exchange of
promises (Baudry, 1992). However, in the particular case of a provision of service, two
features increase the contractual hazards and are sources of transaction costs: the
intangible character of services and the time structure of their effects. Bowen, Jones
(1986) have shown that the intangible character of service increases the uncertainty
affecting the ‘object’ of the transaction (What does the customer purchase?). Because it
is not easy to standardise such services, higher transaction costs are supported by the user
firm compared with goods transactions. These costs are linked to the fact that the user
firm has difficulties evaluating the quality of the service it receives (the output) and the
time and resources (input) necessary to produce the service when it cannot compare it
with another one. Moreover, the user firm is engaged in post-purchase evaluation of the
service when its effects can only be measured in the future. Thus, the existence of this
time lag between the provision of the service and the achievement of its effects entails
higher transaction costs.
(ii) The appropriation of the result of the provision by the producer: The difficulties for the
user firm arise with informational services and can be of two types. On the one hand, the
user firm could have to transfer strategic information to the provider. The risk consists in
the disclosure or the use of this information by the provider for competitive advantage.
On the other hand, the service that is received by the user firm can correspond to its
particular needs though it is not specific, that is the service can be re-deployed to
alternative uses. Thus, the risk consists in the service re-deployment by the provider. The
more the reach of the service is linked to exploring activities, the higher the contractual
hazards (8). In this case, the necessity of protecting strategic information that can be redeployed is a transactional factor that encourages recourse to integrated forms of
organisation.
(8) Analogies are possible with research and development activities when the regime of appropriability is weak
(Teece, 1988, Pisano, 1990).
250
Analysing business services employment — some theoretical and methodological remarks
(iii) Interface and needs specificity: When the object of the transaction corresponds to the
user firm’s specific needs, it has to participate in the provision of service. The ‘relational’
dimension of these personalised provisions allows the user firm to reduce the uncertainty
surrounding the transaction because it has to develop capabilities in the realisation of the
service (production and transfer of information, and capacities necessary to the provision
of service). Thus, the existence of an interface within the user firm can reduce the
uncertainty linked to the evaluation of both service and provider performances. The
consequence of this participation hinges on a co-evaluation of the service by the provider
and the user firm which can be viewed in two complementary perspectives.
First, the co-evaluation is linked to the user firm participation which implies a coresponsibility in the achievement of the result (Gadrey, 1990). In this case, it generates higher
transaction costs associated with more complex negotiation and control mechanisms.
Second, the co-evaluation can be viewed as the result of the user firm’s intention to reduce
uncertainty on services and providers’ performances by investing in the realisation of the coproduction. From this perspective, this is the buyer who realises investments in the aim to
reduce transaction costs.
Thus, business service transactions entail differentiated transaction costs according to the
characteristics of services. From this perspective, the combination of transaction cost theory
and service approaches may explain on the one hand why some services are more prone to be
outsourced than others, and on the other hand why some complex forms of arrangements have
to be found to govern business service transactions. Nevertheless, it is difficult to explain the
growth of the business service sector and the externalisation process by only a decrease of
transaction cost related to service transactions. In particular, one should take into consideration
the growth of needs and the economies of scale in relation to the market size as it is developed
by Stigler (1951). From this point of view, the process of specialisation leads to a vertical
disintegration of industries as the demand for final products (business services) increases.
A complementary point of view consists in explaining the forms of organisation from the
competencies developed in certain kinds of activities in relation to the principal activity of the
firm.
3.4.2.2. Learning processes and service activities
In the competence-based approach, learning processes take part in the definition of the firm’s
boundaries by determining the scope of competencies developed in complementary assets. In
our perspective, service activities represent intermediary and complementary activities
necessary to the user firm’s principal activity. In a ‘knowledge-based economy’ (OECD,
1996), certain kinds of services, and in particular informational services, appear all the more
important as they influence the conditions of the development of the product value (9). In a
(9) See also Philpott in this volume for a discussion of the differentiated jobs created in the service sector by the
weightlesss economy.
251
The job creation potential of the service sector in Europe
general way, informational business services are linked to the conception, the organisation of
the production of the firm and the development of the firm’s strategy in an environment where
the availability of knowledge and information takes part in the development of the firm’s
competitive advantage.
Evolutionary contributions have insisted on the internal learning processes that create the
conditions of development of competencies. The interdependency between firms does not
appear in an explicit way in these works that focus on the specific, dynamic and cumulative
dimensions of intra-organisational learning process. The participation of the user firm to the
service provision leads us to consider various learning processes that influence the
organisation mode of business service activities. In particular, a major difficulty encountered
by the user firm in the provision of informational services consists in the assimilation of the
information and capabilities that have been built during the provision. Cohen, Levinthal
(1990) have shown that a firm’s ‘absorptive capacity,’ that is the ability to evaluate and use
outside knowledge, is a function of the level of prior related knowledge because the firm needs
a knowledge basis in order to assimilate and exploit new knowledge. We encounter the same
necessity of an interface in the user firm able to participate in the provision of informational
service and to exploit the service that it has received. The functions of the interface depend on
the characteristics of the service activities, and in particular on the degree of specificity of the
user firm’s needs (which determines the level of its participation) and on the ‘reach’ of the
service effects (which influences the uncertainty related to the service effects). Thus, the
learning processes have different objects according to the characteristics of the activities. Two
major complementary distinctions can be made.
— First, when the user firm has to participate to the service provision, the interactions
between user and provider generate the conditions for interactive learning process
(Lundvall, 1988, 1992, Cohendet, 1994) which can concern both traditional ‘doing’ and
‘using’ learning, or the ‘search’ activity when the result of the provision is not specified
before its achievement. Thus, according to the degree of the user firm’s need specificity,
learning processes are intra-organisational or interactive.
— Second, the uncertainty related to the ‘reach’ of the service effects influences the way by
which user and provider can achieve the service. The operational services can be realised
on the base of the firm’s experience in the sense that it is possible to improve the provision
by repetition (Levitt, March, 1988) (10), and benefit from increasing returns. This process
can be achieved by interacting if user needs are specific. The exploring services possess
temporal and uncertain effects which have to be learned. Thus, effects are the object of a
‘search’ led internally or by interacting according to the degree of the user firm’s needs
specificity.
(10) This process can lead to a ‘competency trap’ (Levitt, March, 1988) because learning by experience is a
cumulative process which contributes to the risk of specialisation in inferior procedures and routines.
252
Analysing business services employment — some theoretical and methodological remarks
In summary, the propositions made by the theories of the firm depend on the characteristics of
business service activities and on the behaviour of the firm that influences its trajectory and
the things that it is able to make. Considering the interactions between user and provider, we
have seen that the problem of organising these activities is more complex than one of ‘make
or buy’. We now turn to the various forms of externalisation of business service activities.
3.4.3. Proposals for a classification of the forms of externalisation of service activities
The traditional ‘make or buy’ problem concerns the conditions under which firms have an
interest in making a good or service or in purchasing it on the market (Hollard, Ruffieux,
1996). We have considered the limits of applying this problematic to activities which consist
of provision, and in particular, service provision. Nevertheless, the heterogeneity of business
service activities leads us to consider various external forms of organisation along the lines of
‘market-outsourcing-cooperation.’ After presenting the criteria of distinction (Section 3.3.1),
we give some details on the classification according to the characteristics of business service
activities (Section 3.3.2).
3.4.3.1. The criteria
One may use two types of criteria to differentiate the forms of externalisation according to the
level of the participation of the user firm: the originator of the conception and the originator
of the production of service.
— The originator of the conception: Three situations can be considered depending on
whether the conception is made by the provider, the customer or the provider in relation
with the customer. In the first case, we find the standard market transactions for which the
service exists independently of the user firm demand. When the customer conceives the
service, he possesses the ability to define the service and make a ‘purchase on order’ by
specifying the characteristics of the service. These situations correspond to traditional
outsourcing. The last case covers the major part of business service activities. It refers to
situations where the user firm participates in the conception process by defining its needs
and the functional specifications of the service though the provider has the knowledge of
the ‘technical’ possibilities existing. Thus, there is an interactive process in the definition
of the service that has to be produced. We find different definitions of these inter-firm
relationships in the literature: partnership (Baudry, 1995) or ‘speciality outsourcing’
(Morvan, 1991, De Bandt, 1996).
— The originator of the production: As with the previous criteria, the user, the provider or the
provider in relation with the user can perform the production of the service. In the majority
of cases, the provider produces the service but the forms of externalisation depend on the
originator of the conception. We do not consider that the first case corresponds to internal
production within the user firm when it also conceives the service. Nevertheless, it is
possible to consider situations like ‘self-service’ for which the provider conceives the
service though the customer produces it. The very case of ‘co-production’ concerns the
association between provider and user firm in the conception and production of service.
253
The job creation potential of the service sector in Europe
3.4.3.2. The classification
Table 3.2 summarises the various situations of external forms of organisation of business
service activities according to the characteristics of these activities in terms of tangibility and
specificity.
Table 3.2. A classification of externalisation forms of business service activities
Forms of
Kind of
externalisation
service
Conception
Production
Standardised
service
Provider
Provider
High
Low
Self-service
Provider
Customer
Indifferent
Low
Industrialised
service
Customer
Provider
Indifferent
Purchasing
Outsourcing
Level of participation
Features of service
Tangibility level Specificity level
High
Co-conceiving
Co-producing
Personalised
service
Co-produced
service
Provider and
customer
Provider
Provider and
customer
Provider
Provider and
customer
Examples
Low
Low
Low
to medium
Medium
to high
Low
to medium
Low
to medium
High
Installation of
standard software
Database
consulting
Premises
standard cleaning
Tailor-made
packaging
Opinion survey in
a marketing
process
Professional
training
Management
consultancy
3.5. Measurement problems and differentiated job creation potential of
business service externalisation
This section chapter deals with the difficulties of measuring the job creation potential of
externalisation of business service activities. Though externalisation represents a recurring
theme of the specialised press that has removed any doubt of its existence, this phenomena
remains very difficult to consider at a statistical level. In fact, the available statistical field (at
least in France), does not give us a satisfying picture of externalisation at a general level.
However, we share some insights on what could be indirect measures of the process in the
French case (Section 4.1). Finally, we make some hypotheses about the job creation potential
of business service externalisation (Section 4.2).
3.5.1. Measures of business service externalisation and their limits
Externalisation may be defined as a process by which service functions realised previously inhouse are henceforth produced by external providers. So, an ‘externalisation rate’ accounts for
the part of services that are bought by the user firms into the total of intermediary consumption
of services in their production process. Two kinds of measures seem conceivable but they have
254
Analysing business services employment — some theoretical and methodological remarks
a static dimension: the intermediary consumption of services (Section 4.1.1) and the number
of employees affected to ‘service occupations’ (Section 4.1.2).
3.5.1.1. The intermediary consumption of services
This kind of measure deals with purchases of services by user firms. These data are available
in the ‘input-output’ tables produced by national accounts. In France, the entrées-sorties tables
are the only tools which measure the enterprises’ demand for business services (Fontaine,
1987).
Two sets of problems arise with such data. First, the internal consumption of services achieved
by in-house production in user firms is out of the statistical field. So, though it is possible to
have a clear idea of the needs for business services that are bought, the externalisation process
remains an indirectly known phenomenon. Second, the data present a high level of generality
insofar as they deal with ‘business services’ and do not consider the very heterogeneous
services belonging to this category.
3.5.1.2. Employment in ‘service occupations’
This measure of externalisation deals with the number of workers in industrial or tertiary
sectors, classified by occupation. Applied to business services, the measure considers the
number of workers in ‘service occupations’ in the whole economy compared with the identical
number of workers in specialised service enterprises. So, an approximated ‘rate of
externalisation’ could consist in the following ratio: numbers of workers whose profession is
‘A’ employed in the enterprises with ‘A’ as activity, to total numbers of workers whose
profession is ‘A’, whatever the activity of the enterprises.
Concerning this kind of measure, it is necessary to have on the one hand a relatively detailed
breakdown of occupations and activities, and on the other hand a description of these
occupations and activities precise enough to combine them. In France, the lists of activities
and occupations allow such a combination, but the data that intersect occupation and activity
only concern the ‘business service activity’ category as a whole and not its various
components (11).
We thus see that the existing statistical field needs to be developed in order to have measures
of externalisation that allow for comparisons. However, these previous measures are static and
reflect more a ‘rate of outsourcing’ than a ‘rate of externalisation’ (12). We present in Appendix
1 some measurements by combining ‘service occupation’ and the activity of the enterprise in
the French case.
(11) The data are produced by the DARES in the enquête structure des emplois (DARES, 1998).
(12) In order to obtain the latter, it would be necessary to isolate the growth of needs for business services and the
process by which activities are ‘extracted’ from the production process of user firms.
255
The job creation potential of the service sector in Europe
3.5.2. Organisation of service activities and employment relations: Some hypotheses
about measuring job creation potential of business service externalisation
We have seen that it is very difficult to appreciate the job creation potential of business service
externalisation because the available statistical field does not allow us to isolate the growth of
needs and the strict externalisation process. Nevertheless, the analysis of business service
activities has led us to show that the forms of organisation depend on the characteristics of
these activities. In this perspective, we intend in this section to study the working hypothesis
that the job creation potential of business service externalisation is also influenced by the
characteristics of business services and by the way they are conceived and produced. Two
classes of arguments are used.
— First, it is necessary to consider the potentialities of obtaining productivity gains by
externalisation. Several possibilities appear here: (1) the economies of scale reached in the
service sector lead to job losses in this sector on the assumptions that, on the one hand, the
division of labour and the specialisation that it generates allows for productivity gains, and
on the other hand the needs and the demand for these business services are constant; (2)
this previous argument would be compatible with the idea that the industrial sector would
become more competitive after externalisation and such could be the case at the origin of
job creations; (3) in a complementary point of view, the service sector, by generating
productivity gains, also becomes more competitive and generates new demands from the
industrial sector. The consequence would be job creation in the service sector. Finally, the
total effect depends on the weight of job losses linked to productivity gains compared with
these job creations in the service sector.
These speculations assume that there is a possible substitution between internal and external
organisation of business service activities. Moreover, this substitution would be performed
from the only cost differential between these forms of organisation. Nevertheless, we would
argue that for many services, this substitution is not always possible because of the very
characteristics of business services, and it is not always efficient for the firm when we consider
determining factors not to be limited to cost comparisons (13).
In general, the available data on externalisation does not point to dominant behaviours that
substitute one solution for another. For example, in the case of cleaning, which is a service
prone to industrialisation, the ‘outsourcing rates’ reflect the diversity of behaviours in one
country on the one hand, and the divergences of behaviours between European countries on
the other. The rate is 30 % in Denmark, 55 % in France in 1994 and 70 % in the Netherlands (14)
(Panorama de l’industrie communautaire, 1997).
The same conclusions arise with security. The ‘outsourcing rates’ vary greatly between
European countries. Nevertheless, in this case, we have an example of a largely externalised
(13) For such a discussion cf. Benhamou, Gazier in this volume.
(14) The data are detailed in Appendix 2.
256
Analysing business services employment — some theoretical and methodological remarks
service: 60 % in Denmark and Germany, 90 % in France, 98 % in Italy and 66 % in the
Netherlands (Panorama de l’industrie communautaire, 1997).
— Second, the job creation potential of business service externalisation depends on the way
business services are produced and conceived. The study of the heterogeneity of business
services has led us to consider various forms of externalisation within which we find coconception and co-production. These services have a strong relational dimension that
cannot be industrialised unless this dimension is deleted. Thus, we can assume that there
is job creation potential in externalisation both in the service sector and in the industrial
sector: in the service sector because there are few economies of scale potentialities, and in
the industrial sector because it is necessary for the user firm to have an internal interface
to co-conceives and co-produce the service with the provider.
This analysis can be illustrated by the study made by the European Community on
externalisation of business services that shows the extent of the combination of internal and
external forms of organisation for many services (55 % in computer services, 38 % in legal
services, 30 % in engineering, public relations, research and development) according to a
study made by Peat Marwich Mc Lintock for CCE (CCE, 1991).
3.6. Conclusion and policy implications
The development of business services employment is quite rapid today. We showed that part
of it represents a statistical artefact. We underlined the complexity of the organisational
determinants, as regards the employment relationships and the inter-firm relations
surrounding these services. The interplay of these determinants may lead to situations
characterised by low-level equilibria, and possible dual employment structures; for example,
with hidden dependency of subcontractors, low training and low guarantee levels, or with
restricted employment and low output.
Business services also present a high level of heterogeneity that needs to differentiate the
modes of organisation and externalisation behaviour adopted by user firms. We have studied
the ‘make or buy’ dilemma from two complementary perspectives; the transaction cost theory
on the one hand and the competence perspective on the other, the two in combination with
service approaches. Our representation of business service heterogeneity leads us to discuss
the main propositions made by these theories of the firm and to show that these propositions
depend on the characteristics of business service activities (the reach, the intensity of
interactions, the degree of tangibility and the temporalities of services). Moreover, the
problem of externalisation cannot be summarised as a simple transfer of activities and workers
because it can take various forms, some of which imply the participation of the user firm in
the conception and/or the production process.
At a methodological level, it is very difficult to measure the process of externalisation and its
corresponding job creation potential. The existing statistical field does not allow for the
isolation of the transfer of activities between firms and the growth of needs for business
services.
257
The job creation potential of the service sector in Europe
Five main policy consequences stem from our analysis. They correspond indirectly to a
rationalisation and a re-structuring of the sector.
An important new field of intervention and rationalisation emerges when one considers the
connection between labour law and business law: there is a need for new rules, to clarify the
legal structures for collective entrepreneurship, as well as the long-lasting relations between
subcontractors and main firms. The same holds for the status of self-employed workers, who,
with a flow of short-term contracts, are without the normal protection given to salaried
workers. At stake is the visibleness of the real work situation, and the sharing of the risks in
theses activities.
The analysis of the job creation potential of business service externalisation needs more
development of the statistical field on services and on the behaviour of the firm. For instance,
there is little data about the internal production of services by user firms. Moreover, it is
necessary to distinguish the various categories of services and to produce data about these
categories (not about business services as a whole).We also stress the role of more or less
hidden transaction costs in determining which kind of organisation could be chosen, which
have consequences for the producers and the users. Here there is an information challenge
because the management of these costs through flexible arrangements may well induce
instability, risk spillovers and ill-perceived externalities.
One solution frequently adopted in the case of big risks and needed trust and loyalty is to build
networks. However this is not a panacea, because networks need to be open, plentiful, and
equilibrated in order to share risks and disseminate information. There is here a third policy
target: structuring networks brings norms and cooperation.
(1) There is a need for the homogenisation of mobility prospects for workers in the business
service sector, especially through the right to training and access to promotions. In such a
context, the flexibility stemming from the intense competitive pressure of the business
service sector could be more accepted and socially controlled.
(2) Finally, service employment and the job creation potential of externalisation must be
differentiated according to the nature of services. The potential can be assumed to be
higher in the business services corresponding to high-quality jobs for at least two reasons.
First, for these kinds of services, the externalisation process does not correspond to a
simple transfer with a substitution of internal and external solutions. Because these
business services are co-conceived or co-produced, the growth of the service sector
parallels the growth of employment in the user sectors. Moreover, the relational
dimension of these services implies little economies of scale potentialities. Second, there
is a growth of the needs for these services which influences the competitive advantage of
the firms in a ‘service economy’ (15). So, the job creation potential stems as much from
externalisation as from the growth of the needs.
(15) See also Philpott in this volume regarding the ‘weightless economy’.
258
Analysing business services employment — some theoretical and methodological remarks
Appendix 1
Box A1: The main theoretical basis of ‘make or buy’ problem: Contractual and
competence perspectives
— Transaction cost theory (TCT): The contractual perspective
TCT poses the problem of activities organisation as a problem of contracting. The modes of
organisation are defined as ‘governance structure’ which are contractual arrangements in
which transactions are located. The major principle of the model is the following:
‘Transactions, which differ in their attributes, are aligned with governance structures, which
differ in their cost and competence, in a discriminating — mainly, transaction cost
economising way’ (Williamson, 1993, p. 127). Williamson (1985) defines three dimensions
with respect to which transactions are differentiated: asset specificity, uncertainty and
frequency (16). The matching of these three dimensions leads Williamson to determine the
‘efficient governance’, that is the one which minimises transaction costs. The issue is
summarised in Table 2 which refers to Williamson’s works of 1991 and 1993 (17).
Table 2. The organisation of transactions
Investment
Non-specific
Mixed
characteristics
Transaction
frequency
Occasional
Recurrent
Market governance
Trilateral governance
(classical contracting) (neoclassical contracting)
Bilateral governance
(neoclassical contracting)
Idiosyncratic
Unified
governance
(forebearance)
Because of bounded rationality and uncertainty surrounding transactions, contracts are
incomplete and need adaptative and sequential decisions that generate a risk of opportunistic
behaviours from contractors. Thus, when the transaction implies specific investments, the
‘fundamental transformation’ (Williamson, 1985) entails the parties being in a situation of
bilateral dependency which can lead the buyer to the decision of vertical integration.
Resorting to internal organisation avoids repeated opportunistic negotiation linked to
perturbations by appeal to the contractual device of authority and the hierarchical one in the
work relationship. Thus, it is the ex ante choice of internal organisation that allows for the
minimisation of the transaction costs linked to the ex post risk of opportunistic behaviour.
(16) Uncertainty is assumed to be present in an intermediate degree that needs adaptative and sequential decisions.
Frequency (occasional or recurrent) refers to the buyer activity on the market. Finally, asset specificity (nonspecific, mixed, idiosyncratic) refers to investment characteristics.
(17) The classification is slightly different in Williamson (1985); it does not differentiate between the contract law
of unified governance and bilateral governance.
259
The job creation potential of the service sector in Europe
— The competence perspective
In this perspective, the firm is defined as a repository of competencies. Although it is still
difficult to refer to an ‘evolutionary theory of the firm’, many contributions have analysed that
work in the way of such a theoretical building (18). In reference to Nelson and Winter’s book
(1982), competencies are characterised by specificity and their dynamic characters.
Competencies are specific because they are included in organisational routines and they
possess a strong tacit dimension. Moreover, competencies are dynamic because they are
subject to learning (Dosi, Marengo, 1994) which is an organisational, cumulative and
collective process (Teece et al., 1994). For the firm’s boundary problems, this conception has
led to the definition of the firm from its ‘core competencies’ (Teece, 1988) from which it is
possible to delimit its ‘core business’ and to measure its ‘coherence’. A firm’s ‘core business’
encompasses ‘the set of competencies which define its competitive advantage’ (Teece, 1988,
p. 265). They include technological skills, organisational routines and capabilities, and
complementary assets from which the firm can solve organisational and technical problems in
one or more activities. The ‘coherence’ of the firm expresses the fact that ‘their constituent
businesses are related to one another’ (Teece et al., 1994, p. 2). The evolution of core
competencies is constrained by technological trajectories and by the complementary assets the
firm has built in the past. Thus, the limits and the possibilities of evolution of the ‘core
business’ are largely determined by the scope of competencies developed in complementary
assets. Finally, the evolutionary perspective leads to emphasise the importance of the past and
learning processes to determine the firm’s boundaries.
(18) For a presentation of the competence-based approaches, cf. Foss (1993), Hodgson (1998).
260
All activities
Rate of
variation
1987–94
(%)
– 39
– 22
– 5.3
– 0.1
– 16.7
28.9
– 22.5
– 0.2
– 1.2
– 11.3
– 38.2
– 27.9
– 20.1
1990
48.5
36.4
37.3
84.5
30.7
10.5
70.4
68.8
28.8
22
13
39.1
1987
47.9
35.3
37.3
83.5
30.9
9.3
72.4
69.6
29
22.1
16.2
36.5
Industry and agriculture
NAP 15A: U01-U07
NAF 16 EA-EH
43.8
32.6
27.3
81.8
24.6
12.2
64.8
62.2
25.8
17
9.3
34.3
1994
Rate of
variation
1987–94
(%)
– 13.3
– 2.1
57
7.8
– 6.5
24.8
76.6
2.2
54.5
33
– 22.1
0.4
– 1.6
32.9
46
45.7
11.2
35.7
34.6
11.3
9.8
21.3
19.2
19.1
37
1987
31.6
44.9
45.5
9.1
31.2
33.4
17.4
8.9
19.9
17.8
13.2
35.1
1990
42.8
53.2
55.6
11.8
31.8
44
23
9
29.6
22
13.8
48.6
1994
Tertiary without market services /
Business services
NAP 15A:U08-U14 (except U10)
NAF 16: EJ-ER (except EN)
Rate of
variation
1987–94
(%)
– 53.8
– 35.7
28.9
21.8
36.9
– 23.4
– 34.8
56.2
– 0.4
20.8
27.6
– 50.8
– 8.7
19.2
18.7
17.1
5.4
33.4
56.1
16.3
20.6
49.7
58.6
64.8
26.5
1987
20
18.7
17.2
6.4
38.1
56.1
12.3
22.4
51.3
60.2
73.7
25.8
(19) The data only concern the establishments that have more than 20 employees.
We have also calculated the rate of variation of the number of salaried classified by function or speciality. Thus, the number of cleaner has increased by 7.5 % between
1987 and 1994 in all activities, by 27.6 % in the business service sector, it has decreased by 38.2 % in the agricultural and industrial sectors. Moreover, the total
number of salaried has decreased by 13.8 % between 1987 and 1994.
13.3
14.2
17.1
6.4
43.6
43.8
12.3
28.8
44.6
61.1
76.9
17
1990 1994
Market services / Business services
NAP15A: U10
NAF 16: EN
For example, 76.9 % of the employees whose profession is cleaning were salaried in the business service sector in 1994. This is a crude measurement of the
externalisation degree in this activity.
Source: Data produced by INSEE, Structure des emplois au 31 December, INSEE Résultats, Emploi-Revenus, Tableau III (137).
Own calculation (Thèvenot, 1999).
Rate of
variation
1987–94
(%)
Direction
– 33.4
Management and administration
– 15.4
Human resources
29
Organisation of production
1.9
Computing
4.8
Architecture
– 1.8
Geometer
– 13.4
Engineering, technical studies
11.7
Advertising, public relations
11.1
Inquiry and security
16
Cleaning
7.5
Secretarial and translating work
– 23.5
Total
– 10.4
Speciality of service
Sector
Appendix 2: An indirect measure of externalisation: Sector allocation of business service employment by speciality in France, 1987–94
Analysing business services employment — some theoretical and methodological remarks
261
The job creation potential of the service sector in Europe
Appendix 3: Some illustrations of business service externalisation
1.1.1.1.3.1.1.1. The share of computer maintenance contracts
Share of enterprises that use maintenance
contracts according to the activity sector
Share of enterprises that use maintenance
contracts according to their size
Sectors
Industry
Services
Retail trade
Wholesale trade
Transportation
Size
Share ( %)
6 to 9 salaried
50 to 100 salaried
54
64
Share ( %)
53
53
62
72
64
Source: UFB-Locabail, in Xerfi, Secteurs 700–1997, Série 701, Services aux entreprises Vol. 2,
p. 108.
The outsourcing market of cleaning ( %)
Country
1989
1991
1992
1993
1994
Belgium
Denmark
Germany
Spain
France
47
23
—
37
43
55
25
65
55
46
55
30
65
60
48
55
30
65
60
50
55
30
65
60
55
Netherlands
Portugal
United Kingdom
55
45
20
58
45
30
63
60
35
67
63
39
70
65
40
Source: Fédération Européenne du Nettoyage Industriel (FENI), Panorama de l’industrie
communautaire 1997.
Outsourcing rate of security, 1996, (%)
Country
Belgium
Denmark
Germany
Greece
Spain
France
Ireland
Italy
Luxembourg
Netherlands
Austria
Portugal
United Kingdom
Outsourcing rate
78
60
60
70
99
90
82
98
70
66
75
70
67
Source: Confédération Européenne des Services de Sécurité (CoESS), Panorama de l’industrie
communautaire.
262
Analysing business services employment — some theoretical and methodological remarks
Externalisation level of business services in the European Community, 1988
Origin of service
Exclusively external
Exclusively internal
Combination of
external and internal
Sector
Engineering and linked
Management consultancy
Advertising
Public relations
Computer services
Research and development
Legal services
Operational services
Country
Germany
France
Italy
Netherlands
United Kingdom
Size (employment)
0-50
51-500
501-1000
1001-5000
5000+
56
35
49
11
22
12
41
58
14
37
24
59
26
58
21
22
30
28
27
30
55
30
38
20
32
56
47
39
34
47
30
33
18
23
21
14
18
43
43
37
56
38
39
37
44
22
29
35
33
19
22
33
26
30
Source: Panorama de l’Industrie communautaire, 1991–92, Study of Peat Marwick Mc Lintock, ‘The
costs of non-Europe for business services’, CCE, (1988).
263
Part III:
New perspectives of the service economy
Chapter 1
Weightlessness and the political economy of
service sector job creation
John Philpott
1.1. Introduction
References to the ‘new economy’ are nowadays commonplace in books, political speeches
and official policy documents. ‘New’ in this context means far more than simply ‘modern’.
The underlying theme is that developed economies are increasingly knowledge dependent,
founded on information and communications technology, and ever more geared to the
production of services rather than primary and manufactured goods.
The new economy is thus a super buzz-phrase encompassing others such as ‘the knowledge
economy’, ‘the information economy’, ‘the digital economy’, ‘the service economy’ and —
most noteworthy of all — the ‘weightless’ (or ‘dematerialised’) economy. The latter term
stems from the observation that the real value of what developed economies produce now far
exceeds the physical weight of output. The initial observation is often attributed to Alan
Greenspan, Chairman of the United States Federal Reserve. He noted in 1996 that whereas the
weight of the United States output in the 1990s was little different from the 1940s the real
value of output was three times higher.
The reasons for this are fairly straightforward. The service sector accounts for a larger
proportion of developed economy gross domestic product (GDP) than a generation ago (the
OECD average is now around 70 %). Technological improvements meanwhile have altered
the physical attributes of manufactured goods, thereby making them lighter (miniaturisation
being the most obvious example). Moreover, far more of the value of goods and services
resides in the higher quality that richer and more sophisticated consumers demand of them,
notably in the form of design and specification.
267
The job creation potential of the service sector in Europe
The labour market corollary is that the new economy is only capable of creating weightless
work. Fewer workers are employed in the production of tangible goods and those that remain
are increasingly knowledge workers — needed to improve the intangible qualities of products
— rather than either unskilled muscle workers or routine skilled or unskilled manual
operatives. This in turn leaves the service sector — which produces intangibles — as the sole
source of net employment growth.
Economic weightlessness is a classic ‘oh yeah!’ fact — everyone is aware of the phenomenon
in their everyday lives as workers or consumers but still surprised and intrigued when it is
pointed out to them. The problem with such facts is that once heard the novelty soon wears off
and they tend to be treated as statements of the blindingly obvious. With intangibles
accounting for a greater share of what is produced, technology advancing, and much of the
value of goods and services itself intangible, it is perhaps not surprising that developed
economies have grown without gaining weight and offer new forms of light employment.
For many economists the temptation is thus to dismiss references to the weightless economy
as little more than trendy terminology for the technological change and deindustrialisation that
has characterised developed economies since the late 1960s — the implications of which have
been the subject of debate for many years. However, to do so is short-sighted. As this chapter
demonstrates, an examination of the issues highlighted in the literature on the new economy
and economic weightlessness opens up fresh perspectives on discussion of the job creation
potential of the service sector in Europe and related policy issues.
For example, the new economy literature challenges apocalyptic ‘end of work’ scenarios of
economic development. Work is becoming weightless because the types of jobs being created
are different, not because there is less need for workers. Indeed, affluent service-based
economies have the potential to create more jobs, not fewer. In view of this the fact that
deindustrialisation has coincided with mass unemployment in the EU during the past two
decades is paradoxical. The explanation lies partly in relatively slow rates of output growth.
However, more important, policy failures have prevented many EU Member States from
maximising the employment potential of the service sector, as evidenced by the EU’s service
sector ‘job deficit’ when compared with the United States.
Employment in services accounts for around 55 % of the working age population in the United
States but only 39 % in the EU (the United Kingdom figure is around 50 %). Whilst this deficit
may be considered bad news when viewed as a measure of relative labour market
performance, it is good news insofar as it indicates unfulfilled potential for service sector
employment in the EU. Making use of this potential could help the EU move back toward full
employment — which explains the European Commission’s interest in the current Resnet
project.
Yet mention of the need to create more service sector jobs is associated with unease in many
EU Member States where for most of the post-war era manufacturing has been considered the
mainstay of ‘good’ employment. Service sector jobs, by contrast, are often portrayed as being
mainly poor quality low-paid jobs, summed up in the condescending phrase ‘McJobs’. A focus
268
Weightlessness and the political economy of service sector job creation
on the new economy literature is also useful in this respect, because it both highlights what is
required to create weightless service sector jobs and demonstrates that, contrary to the popular
stereotype, such jobs provide high as well as low quality employment, much of it knowledgebased. However, this literature makes no promise of an automatic guarantee of a jobs Nirvana.
Weightlessness is not the same as painlessness; there are winners and losers in the new
knowledge-driven economy, an outcome that poses serious policy questions.
Section 1.2 of this chapter discusses the causes and consequences of weightless growth and
assesses what this means for employment and the labour market. The chapter draws primarily
upon the United Kingdom literature and data but should be of interest to the wider EU
audience (1). Section 1.3 then considers the related broad policy implications and sets out
criteria that policy-makers might wish to take into account when deciding upon the relative
merits of policies designed to create service jobs in the private, public and social economy
sectors.
1.2. Weightless growth and employment
Many conventional accounts of deindustrialisation offer a fairly pessimistic view of the effects
of structural change on employment. Improvements in productivity combined with shifts in
demand as incomes grow reduce the number of jobs in primary and manufacturing industries,
with labour transferring into service sectors where productivity is lower.
Technological change plays a large part in this story — often in association with globalised
trade — with the corollary that productivity gains will eventually spread into services leaving
only service jobs that are immune to technology as the mainstay of employment. Whilst this
conclusion is far removed from the simplistic ‘end of work’ scenarios of economic
development peddled by crank futurologists, the picture that emerges is nonetheless rather
sobering. Not so much the weightless economy as the anorexic economy.
Accounts of the new economy, by contrast, paint a more optimistic vista of job and incomerich weightlessness. Human creativity and skill is presented as being vital to making the most
of knowledge-based technology, while ‘the human touch’ is essential to the provision of a
wide variety of personalised services, both high-tech and low-tech. This section looks at the
central role played by knowledge in the weightless economy and how this affects
developments in the labour market.
(1) This section draws heavily upon a number of key sources. Coyle, D. (1997), UK Department of Trade &
Industry (1998), Quah, D. (1999) and Leadbeater, C. (1999).
269
The job creation potential of the service sector in Europe
1.2.1. The knowledge economy
Knowledge is generally considered the main driver of the new economy ( 2). Of key
importance are four main types of knowledge activity, all of which are intrinsically light and
intangible:
• Information and communications technology, including the Internet.
• Intellectual property, including not only patents and copyrights but more broadly brand
names, trademarks, advertising, financial and consulting services, health care and
education.
• Electronic libraries and databases, including new media, video entertainment and
broadcasting.
• Biotechnology, which include carbon-based libraries and databases as well as
pharmaceuticals.
Knowledge is not only an essential input to these activities but also the prime output, either in
the form of knowledge products or as systems for managing and transmitting knowledge. The
various activities are centred on specific knowledge-based sectors of the economy. These in
turn can be classified as high-tech industries (aerospace, computer and office equipment,
communications, equipment and pharmaceuticals) and knowledge-based services (telecoms,
computer and information services, finance and insurance, education and health). Together,
the knowledge-based sector accounts for more than 50 % of OECD-wide business sector value
added — up from 45 % in the mid-1980s (3). Investment in knowledge (research and
development, computer software and public and private spending on education) represents
about 10 % of OECD-wide GDP. However, the influence of the knowledge economy extends
far beyond the boundaries of these sectors.
1.2.2. Knowledge as an engine of growth
Knowledge products and services can be applied throughout economies and indeed across
geographical borders. As a result, information and communications technology (ICT), for
example, has been likened to steam power and electrification in previous eras — the new
knowledge-based technological mode affecting not just a few products and services but
practically every product and service (4). Improvements in ICT have both reduced the cost of
computerisation and made information systems less expensive and more effective. Between
1992 and 1999 alone the United States quality adjusted computer hardware price index shows
a 40 % reduction for large computers and a 90 % fall in the case of personal computers.
(2) Quah (1999) op. cit.
(3) OECD (1999).
(4) Freeman, C. (1996).
270
Weightlessness and the political economy of service sector job creation
Examples such as these make it easy to understand why the generation and application of
knowledge-based technology is said to be the major growth engine of the new economy.
However, equally important is the fact that all of these knowledge products and systems can
be used frequently and simultaneously by business and consumers without depleting the initial
stock. This quality, referred to as ‘infinite expansibility’, opens up even greater potential for
economic growth albeit creating a dilemma for the producers of knowledge products and
society at large (5).
A key characteristic of knowledge products is that the marginal cost of production falls toward
zero. Given that in competitive markets product prices are driven down to marginal cost,
companies may be reluctant to invest in knowledge products unless they can gain sufficient
market power to enable them to price above marginal cost. In some cases this market power
is sanctioned by the state in the form of patents — an acknowledgement that there is a tradeoff between the economic and social advantages of making the fruits of knowledge available
at the lowest possible price and the need for incentives to produce knowledge products in the
first place. In general, however, governments are keen to foster product market condition. As
a result, companies seek to gain market power by clearly distinguishing their products from
those of competitors — hence the growing importance of brand names and advertising in the
new economy.
1.2.3. Explaining the ‘productivity paradox’
Despite the obvious growth potential of knowledge-based technology, commentators have
often questioned the degree to which the developed economies are experiencing a
technological revolution. Best known is Solow’s ‘productivity paradox’ — summed up in the
Nobel Laureate’s much quoted comment that ‘you can see computers everywhere but in the
productivity statistics’. Growth in productivity and output were faster in the old industrial era
of the 1950s and 1960s than in the 1980s and 1990s. Several explanations have been suggested
for this, ranging from greater difficulty of measuring output in service economies to the slow
diffusion of technology.
The most important explanations refer to problems surrounding the effective transfer and
absorption of knowledge technologies. Knowledge is truly transferable only if formalised and
codified. However, a lot of the knowledge actually being applied in developed economies is
of an informal or tacit kind. Whilst this can be a powerful source of competitive advantage for
individual companies and/or nations, its power as a general driver of improvements in
productivity and economic growth is more limited. Perhaps the most persuasive explanation,
however, is that which relates to what is called the ‘inherent unknowability’ of knowledge
products and systems (6).
(5) Quah (1999) op. cit.
(6) Quah (1999) op. cit.
271
The job creation potential of the service sector in Europe
Put simply, this means that, when confronted with new forms of knowledge, businesses and
consumers are often uncertain about how to use them and perhaps even suspicious of them. It
thus takes time before the potential of knowledge can be fully utilised. It has been suggested
that the time lag could amount to more than a generation. However, there is optimism that
recent signs of a productivity improvement in the United States — productivity is currently
rising at a rate of 3 % per annum compared with just 1 % a decade ago — indicate that the
productivity paradox is beginning to be resolved.
In many societies the ‘tried and trusted’ is preferred to the new, while more generally the ‘pretech’ generation are slower to adapt to change than the young. In October 1999, for example,
the United Kingdom Prime Minister, Tony Blair, had his first tutorial in the use of the Internet,
commenting in the process that his children were putting him to shame in the computer stakes.
In a similar vein, consumers may initially suffer from ‘technophobia’ when confronted by new
knowledge-rich products that limit demand. In the 1980s, this was evident with regard to the
computer-based products but consumers have since been on a fast-rising learning curve, as
evidenced by the number of ‘wired’ households now making use of the Internet. At present,
the biggest source of technophobia is seen in consumer wariness of genetically modified foods
which may limit the growth potential of this aspect of the biotechnology industry.
Despite constant talk of technological revolutions most developed economies thus remain a
mix of the new and the old with many businesses, workers and consumers blissfully attuned
to the latter. Some British commentators, for example, argue that too many companies remain
wedded to low quality products involving little knowledge input because high-quality product
strategies require substantial investment and radical changes in corporate organisation (7).
This of course gives rise to externality problems since what makes short-run business sense
for these companies is not in the long-run interest of society as a whole.
This kind of cultural or institutional aspect to the speed with which technology is applied and
spread throughout the economy is also relevant to broader debates about the role of
technological change in the process of economic growth. Whereas technological progress has
traditionally been viewed as ‘manna from heaven’, it is increasingly recognised that the
underlying growth rate depends critically on the way in which market and social institutions
are geared to the generation and utilisation of knowledge (8).
The latter view stems from ‘post-neoclassical endogenous growth theory’ — once famously
referred to by the United Kingdom Chancellor of the Exchequer, Gordon Brown, in a speech
much lampooned by the media. This rather esoteric sounding theory — which boils down to
the conclusion that institutions matter and that policy can make a difference, especially by
encouraging investment in knowledge skills and technology — nonetheless underpins much
of the current United Kingdom Government’s approach to making the most of the new
economy so as to raise productivity, growth and employment.
(7) Keep, E. and Mayhew, K. (1998).
(8) Romer, P. M. (1990).
272
Weightlessness and the political economy of service sector job creation
1.2.4. Weightlessness and employment
Debate surrounding the impact of knowledge-based technologies on productivity and growth
spill over into debates about ‘the future of work’. Some commentators talk of revolutionary
change in the labour market while others temper this by highlighting the strong degree of
continuity that exists (9). Since as already mentioned, economies are an amalgam of the old
and the new, evidence can be used to support either perspective. United Kingdom data, for
example, show little change in average job tenure since the 1970s casting doubt on the
common assertion that employment is becoming more unstable and that the ‘job for life’ is a
thing of the past. By the same token, however, there have been some significant changes in the
industries in which people work and the types of jobs being performed which means that the
labour market at the start of the third millennium is very different from that of a generation
ago.
At first sight the observed pattern of employment change in the United Kingdom and most
OECD economies would appear consistent with the standard account of structural change.
Jobs have been lost from all parts of manufacturing — light high-tech knowledge-intensive
‘sunrise’ industries as well as traditional heavy ‘sunset’ ones — with net job creation confined
to the service sectors. Manufacturing now accounts for fewer than one in five United Kingdom
jobs, down from one in three thirty years ago.
The decline in manufacturing employment is explained partly by productivity gains resulting
from technological change and partly from productivity gains inspired by the need to reduce
costs in the face of competition from non-OECD countries. In the United Kingdom, periodic
bouts of job shedding from manufacturing since the 1980s have also been associated with the
overvaluation of sterling on currency markets.
The relative contribution of technology and trade (or globalisation) to the loss of
manufacturing jobs has been a hotly contested topic for several years (10). The jury is still out
on this issue albeit the consensus favours the technology story (or some link between trade and
technology). However, what is apparent is that there is no simple relationship between
technology, productivity and employment growth in the service sector.
Although productivity growth in some knowledge-based service sectors like
telecommunications has resulted in a fall in employment, knowledge-based service sectors as
a whole have experienced job growth alongside productivity gains in response to rapidly
growing demand from businesses and consumers. The index of employment in knowledgebased industries in the OECD thus rose from 100 to 118 between the mid-1980s and mid1990s (11). In the United Kingdom employment in knowledge-based services has grown faster
than in areas like retailing, the largest increases being in finance, insurance, real estate and
(9) Meadows, P. (1999).
(10) Freeman, R. (1996).
(11) DTI op. cit.
273
The job creation potential of the service sector in Europe
business services. The share of the latter in total UK employment has risen from around 10 %
to around 20 % since the early 1970s (12).
1.2.5. The new micro-service economy
These sectoral shifts show only one dimension of the changing profile of jobs in the new
economy. Even more revealing are occupational data (13). These highlight three clear trends
since the early 1980s. First, a marked decline in manual employment, skilled and unskilled;
the occupational dimension of the fall in manufacturing jobs. The share of skilled manuals in
total employment has fallen from 18 to 13 %. Secondly, a rise in skilled employment of people
performing managerial, professional and technical jobs. The share of professional workers has
risen from 7 to 13 %. And thirdly a rise in essentially low-skilled employment performed by
a group classified as ‘personal and protective’ workers — their share of total employment has
doubled to 14 %.
Manufacturing accounts for some of the growth in skilled employment; net job loss in the
sector has coincided with increased demand for knowledge workers — engineers, scientists,
designers and the like. However, the major determinant of employment growth, whether highor low-skilled jobs, is increased demand from businesses and consumers for a wide variety of
professional and/or personalised services. These range from high-tech knowledge-based
services to ‘high touch’ services. Collectively these suppliers comprise a new broad subsector
of micro-service providers, the emergence of which can be traced to developments in both the
corporate and household sectors of the economy.
1.2.6. Outsourcing and networking
Within the corporate sector increased demand for micro-services stems from a tendency for
large companies to focus on a core of essential functions and buy-in non-core services rather
than produce them in-house and incur all the necessary overhead costs. This practice known
as ‘outsourcing’ challenges the concept, initially formulated by Coase, of companies as large
virtually integrated corporate citadels operating on the principle of command and control.
Bought-in services commonly include catering and cleaning as well as long-standing
professional services, like law and accountancy. However, businesses are also buying-in new
professional services, such as software consultancy and web site design. Services of all these
kinds are supplied either by specialist contracting companies of varying size, or by selfemployed contractors or consultants (often known as ‘portfolio workers’ if they provide
services to a number of business clients).
It is important not to confuse outsourcing with the common notion that companies are
reducing their core permanent workforce to the bone and relying instead on a periphery of
(12) Institute of Employment Research (1999).
(13) Ibid.
274
Weightlessness and the political economy of service sector job creation
temporary workers who can be taken on and laid off at will and incur fewer staff overhead
costs. The employment effects of these two distinct practices are very different. Whereas
outsourcing appears to be a fairly widespread practice in the United Kingdom there is little
evidence of a marked trend to toward adoption of the core-periphery workplace model (14).
Temporary employment accounts for only 6 % of the employed workforce in the United
Kingdom, little different from a decade ago.
In statistical terms it is sometimes argued that outsourcing of services by manufacturing
companies exaggerates both the decline in manufacturing employment and the rise in service
employment. As a result differences between economies in the degree of outsourcing can
distort international comparisons of the relative shares of manufacturing and service
employment. A related but broader argument is that outsourcing does not lead to net
employment generation in the economy since outsourced jobs provided by micro-service
enterprises or the self-employed would in any event exist within large companies. Anecdotal
evidence abounds of individuals leaving large companies as employees on a Friday only to
turn up the following Monday as self-employed consultants (which explains current
controversy in the United Kingdom over the tax treatment of consultants) (15). Moreover, if
contracting firms can supply outsourced services more efficiently outsourcing might even
reduce employment.
Arguments such as these are often associated with the ‘manufacturing matters’ view. This
questions whether economies can be truly serviced based on the grounds that manufacturing
is the prime source of income to be spent on services. In this respect it is necessary of course
to distinguish the contribution of manufacturing to output from its contribution to
employment; even a healthy manufacturing sector is no longer a source of net employment
generation. However, debates over the relative primacy of manufacturing or services anyway
appear rather stale in the new economy where knowledge is the wellspring of growth — a
conclusion that itself has some bearing on the relationship between outsourcing and
employment.
Although the practice of outsourcing is normally associated with low-tech services such as
catering and cleaning, as already mentioned it also encompasses the buying-in of knowledge
skills from ‘ideas’ contractors, like computer software or design consultants. It is now said that
companies make use of such external knowledge skills by utilising ICT to establish strategic
or creative networks — a phenomenon referred to as ‘virtual integration’ or ‘wired world’ (16).
Such networks or joint ventures can result in long-standing collaborations or fluid project-byproject relationships that can be established and disestablished (what might be called ‘velcro’
or ‘Lego’ ventures). Networks can also involve partners all around the world using ICT to the
(14) See, Philpott, J. (1999).
(15) Burton, J. (1999).
(16) Leadbeater op. cit. See also DTI (1999).
275
The job creation potential of the service sector in Europe
full, albeit in practice a lot of collaboration involves business clustering. However, either way
a strong degree of trust is required to make them work.
Specialist small businesses or self-employed consultants are considered crucial to the
effectiveness of such networks since they impart flexibility, flair and creativity of a kind that
larger corporations are not necessarily able to foster themselves. This is very different from
low-tech service outsourcing and much more important to the overall growth and employment
potential of the economy. Knowledge outsourcing should not therefore be considered as
simply reallocation of labour from one part of the economy to another.
The suggestion that networking will inevitably provide a major boost to employment at the
knowledge end of the micro-service economy is itself questioned, however. For reasons
already discussed, some commentators consider ‘wired world’ an idealised typology in an
economy where many businesses still seek to operate in low-tech markets. Consequently,
although knowledge-based networking might be an identifiable trend it will not necessarily
become a dominant corporate form. Moreover, other commentators believe that even if more
businesses do gear-up for the knowledge economy the large corporation will stage a
renaissance (17).
The idea here is that investors will place ever greater weight on the value of companies’ human
capital and knowledge talent, encouraging businesses to protect knowledge assets by
developing them in-house rather than buying them in. Likewise, companies will seek to secure
a comparative advantage in product markets by developing tacit knowledge within corporate
boundaries rather than networking on the basis of transfer of codified knowledge.
In this context ICT networking will still be conducted but intra-firm rather than inter-firm. If
so, the rise of the portfolio workers could turn out to be a short-lived phenomenon of the 1980s
and early 1990s rather than a long-run trend, with the next decade witnessing an increase in
teleworking involving intra-networked employees. This might be one reason why selfemployment has fallen in the United Kingdom in the 1990s having experienced a meteoric rise
in the 1980s (self-employment currently accounts for 12 % of total United Kingdom
employment, only a shade higher than in the mid-1980s). There is, however, no evidence as
yet of a marked rise in teleworking despite the best efforts of commentators to talk-up the
phenomenon (estimates of the share of teleworking in total United Kingdom employment
range from 1 to 5 % depending on the definitions used).
1.2.7. Households
Demand for professional and personalised services within the household sector has grown out
of a combination of two trends: rising real incomes and an increase in the proportion of dual
or multi-earner households.
(17) Ibid.
276
Weightlessness and the political economy of service sector job creation
In recent years considerable public and policy attention has been given to workless
households. However, this has been merely one aspect of a growing polarisation of work. At
the other end of the scale there has been a rise in the proportion of people living in ‘job rich’
multi-earner households — up from 70 % in the mid-1970s to 80 % in the mid-1990s (18). The
latter development reflects an interaction between the types of jobs on offer in the new
economy and social change which has increased female participation in the labour market.
Women gaining professional qualifications have taken advantage of rising demand for
workers with problem-solving and creative skills. More generally women have benefited from
the shift into service sector employment which has offered greater opportunity for part-time
work, enabling women to combine jobs with domestic roles. However, all of the increase in
female working since the late 1970s has been in households where the partner also works.
The obvious difference between dual earner households and the traditional sole earner
households of the 1960s and 1970s is that they have higher incomes and also value time more
highly. This means that they are more likely to purchase goods and services that reduce time
spent on household chores (everything from convenience foods to fast food and domestic
help) and higher quality goods and services that can be enjoyed during whatever spare leisure
time they have available (whether a visit to a top class restaurant, the local multi-plex cinema
or a cosmetic makeover).
As a result, the types of personalised services that are in demand can be quite wide ranging,
with professional services at one end of the spectrum and services like house cleaning or dog
walking at the other. The mid-range includes services combining high-tech with high touch —
music tutors, chiropractors, interior designers and homeopaths, etc. — offering a personal
service involving some specialist skill or knowledge (albeit based on qualifications with
varying degrees of professional standing). Unfortunately, this is a very under-researched area
and it is not known to what extent these services are supplied by sole traders as opposed to
somewhat larger businesses. What is likely, however, is that demand for such services will be
cyclical.
The combination of high income and scarce leisure time is most marked in dual earner
professional households (who may have earned income in excess of GBP 70 000 per
annum) (19). These are not only capable of purchasing a large volume of services but also the
widest variety of high and low value added services. Because such households tend to be rich
in knowledge and information as well as income this can result in increased demand for
services such as Internet shopping and banking. This raises the potential for job creation in the
growing e-commerce sector, although the overall impact on employment will depend on the
extent to which wired shopping opens up new markets or simply substitutes for existing jobs
in retail or financial services.
(18) Gregg, P. and Wadsworth, J. (1996).
(19) Hutton, W. (1999).
277
The job creation potential of the service sector in Europe
The use of information technology in this way to some extent de-personalises the provision of
these services since it breaks the physical or geographical link between purchaser and
provider. Such services occupy cyberspace rather than physical space which means that jobs
created as a result of Internet shopping and banking are not necessarily located in traditional
retail centres. However, this is still not the case where the human touch, rather than the
disembodied human voice or electronic impulse, is essential to the service provided.
In affluent areas with agglomerations of high-income households, localised expenditure on
personalised services can thus be considerable. Some estimates suggest that the average dual
income household in London spends GBP 5 000 each year on services of this kind (which may
be an underestimate because of informal tax dodging in the supply of personalised
services) (20). Given that these professional households comprise the knowledge elite of the
new economy it is possible to observe a link between the rise in the numbers of people living
in knowledge occupations and those engaged in the personal and protective occupations.
1.2.8. Weightless worries
Despite the potential for growth in employment of all kinds in the weightless economy the fact
that there are losers as well as winners nonetheless raises concern about ongoing
developments in the labour market (21).
It is important to appreciate that the knowledge economy has posed problems for a wide
variety of workers, not just manual workers shed from manufacturing. For example, certain
categories of manufacturing workers have been displaced as companies have downsized and
restructured managerial hierarchies. Whatever the impact of the knowledge-based technology
on the Coasian corporate citadel it is evident that Tayloristic ‘top-down’ management
practices have been eroding. More companies have introduced flatter management structures
and operate by way of team working and internal networks (22).
Banks and financial institutions meanwhile have been rationalising high street branches and
instead establishing more call centres, resulting in net hiring of call centre staff but a reduction
in more traditional forms of clerical work. Call centres have been identified as the fastest
growing source of new jobs in the United Kingdom, with between 150 000 and 350 000
employees at present according to available estimates. Call centres can operate at lower cost
and, most important of all, around-the-clock. There is concern that this is leading to the
emergence of a ‘24 hour society’ with adverse consequences for workers and their families
albeit criticism that call centres offer only low pay and poor working conditions and are
plagued by high labour turnover is not always justified (23).
(20)
(21)
(22)
(23)
278
London Evening Standard, 20 September 1999.
Coyle op. cit.
See Cully, M. et al. (1999) Britain at Work. Routledge.
Incomes Data Services (1999). See also Mitial (1999).
Weightlessness and the political economy of service sector job creation
What is significant is the fact that call centres do not have to be located close to the consumer
in order to provide services. Net job creation in call centres may therefore be of little comfort
to workers in financial service sectors made redundant by high street branch closures.
Moreover, the fact that proximity to the consumer is no longer a pre-requisite of service
provision in the information age raises the spectre of ‘service globalisation’. This is the
possibility that information-based service sector jobs might be lost as a result of employers
transferring data processing tasks etc. to countries with lower labour costs. Within the United
Kingdom, for example, call centre jobs tend to be created in low-wage areas where
unemployment is high. While such jobs are normally welcomed the fear is that, should wage
levels rise, jobs will move away.
Knowledge-driven restructuring of the various kinds described above, combined with
justified or unjustified worries about globalisation, almost certainly underlies a heightened
sense of job insecurity in United Kingdom society. It is currently fashionable for United
Kingdom commentators to argue that increased job insecurity is a myth, pointing to evidence
to show that the proportion of workers experiencing concern about losing jobs has not
changed since the mid-1980s. The inference is that the sense of insecurity has been whipped
up by opportunist politicians and a scaremongering media. It should be remembered, however,
that United Kingdom unemployment in the mid-1980s was double the rate in the 1990s. A fall
in job insecurity might therefore have been expected, rather than constancy (24). This is likely
to reflect the fact that even in a relatively buoyant labour market workers at the sharp end of
structural change may feel insecure.
For example, according to opinion polls around half the United Kingdom workforce fear
being left behind with regard to developments in information technology. Such fears may well
be heightened amongst workers with an older knowledge ‘vintage’ — especially older men, a
group that have experienced a substantial reduction in average job tenure since the 1970s.
Job fears might also stem from the emerging pattern of employment change by occupation as
discussed in section 1.5. There are more well paid knowledge jobs and more less well paid
‘personal and protective’ service jobs but fewer ‘mid-range’ jobs of a kind that were once
plentiful in manufacturing. Moreover, the pay gap between knowledge workers and the
remainder of the workforce has grown as the pay of the knowledge elite has accelerated. In
the United Kingdom, the ratio of the top to the bottom of the male earnings distribution rose
from 2.5 to 3.1 between the late 1970s and the mid-1990s.
Higher pay at the top of the distribution is observed in its most extreme form in the ‘winnertakes-all’ phenomenon. This arises where information technology and/or the growth in
international markets enables highly talented individuals to command astronomic salaries far
in excess of the average for their profession. For example, cheap and easy replication of music
(24) Felstead, A. et al. (1998).
279
The job creation potential of the service sector in Europe
and videos opens up mass markets for the world’s top entertainers. Cross border competition
likewise drives up the pay of company executives.
What is even more interesting, however, is that a kind of ‘winner-takes-all’ effect is seen
within all occupations, that is, the pay gap is not only widening between different groups of
workers but also within different groups of workers. Whilst the reasons for this are not entirely
clear, it would appear that in the new economy workers are being rewarded for that ‘extra
something’ — the undefinable human element people bring to the labour market whatever
their position in the job hierarchy.
A mid-range worker is likely to view the ‘hollowing-out’ of the labour market and the
resulting inequality with some trepidation, especially if unprepared for the knowledge
economy. Workers in this situation who worry about their employability may thus feel
insecure not because of fear of imminent job loss but instead because of what might happen if
they were to lose their job. Those unable to ‘trade-up’ in the new economy may worry that they
will have to trade-down in order to stay in work and accept lower pay (United Kingdom
workers who lose their jobs typically find that re-entry jobs on average pay around 20 % less).
1.2.9. New economy, third way
The shifts in labour demand that give rise to weightless worries are at the root of the structural
unemployment and/or growing inequality that one way or another bedevil all the EU Member
States as well as other developed economies. Within Europe this gives rise to a rather polarised
debate founded on the assumption that societies inevitably face an adverse trade-off between
‘low pay or no pay’.
On one side, there are those who advocate worksharing or a large dose of citizens income
transfer to shield the losers from social exclusion. On the other side, there are those who call
for deregulated labour markets and radical welfare reform to ensure that the losers are priced
into jobs.
Those in-between include traditional Keynesians who attempt to side-step the trade-off by
arguing that Europe’s employment problems have largely macroeconomic causes and thus
require aggregate demand side solutions.
The in-betweens also include those who acknowledge the existence of structural problems but
argue that the notion of a ‘low pay or no pay’ trade-off is far too simplistic. On the basis of an
observation of the way in which relative demand is shifting in the new economy the latter
conclude instead that modern developed economies can enjoy high rates of growth and
employment and, moreover, are capable of creating well paid as well as low paid jobs.
The corollary is that this positive outcome will not occur automatically in the absence of an
appropriate policy response designed to create conditions of macroeconomic stability, foster
improvements in productivity and address structural labour market problems.
280
Weightlessness and the political economy of service sector job creation
This type of conclusion has much in common with so-called Third Way thinking; not
surprisingly there is a strong Third Way element in much analysis of the new economy.
However, the consensus in favour of this kind of policy approach extends well beyond those
normally associated with the Third Way. The remainder of this chapter thus considers the mix
of employment polices likely to be most conducive to the creation of jobs in the weightless
economy.
1.3. Employment policy in the weightless economy
In terms of micro and employment policies the response to the economic challenges posed in
the new economy requires measures that, (a) increase the supply of knowledge workers to all
sectors of the economy, (b) facilitate the development of market and social institutions
conducive to the best use of knowledge, (c) foster entrepreneurship and the creation of service
sector jobs and (d) adjust tax and benefits systems to support low productivity workers in
employment rather than in unemployment or inactivity.
These are necessary conditions for the establishment of full employment and will result in net
job creation if implemented in the context of macroeconomic and industrial policies that
ensure macroeconomic stability and encourage improvements in productivity.
Given the structural changes affecting developed economies net job creation will be confined
to the service sector. This section briefly examines the general conditions for service sector
job creation in the new economy and then sets out criteria against which to judge the relative
merits of job creation in the private, public and social economy sectors.
1.3.1. General conditions for service sector job creation
The most basic requirements for job creation in the new economy are an increase in the supply
of knowledge workers, flexible internal workplace structures that can make the best use of
knowledge workers, competitive product and capital markets, and employment friendly tax
and benefit systems.
Variations on this basic theme — with adjustments made to suit individual ideological whim
— can be seen in policy prescriptions flowing from the G8 and the OECD, as well as in the
recommendations of the high-level group established by the European Commission to study
the economic and social implications of industrial change (25). They of course provide the
essence of the four pillars of the European employment strategy and the related employment
guidelines. The guidelines, therefore, provide a useful reference point for discussion of the
conditions needed for job creation.
(25) See OECD (1992). See also European Commission (1998).
281
The job creation potential of the service sector in Europe
Increasing the supply of knowledge workers will not only meet growing demand for such
workers but also enable a faster rate of growth in productivity and output, thereby creating
higher incomes and increasing demand for employment as a whole. To achieve this, education
and lifelong learning policies need to instil the ability to use and impart knowledge. This
implies the teaching of information management, personal networking skills and problem
solving skills. In other words, education and initial training should focus primarily on the
provision of generic skills, key skills and ‘soft’ (that is, personality) skills rather than narrow
specific academic or vocational qualifications (26).
In the new economy workers benefit most from transferable skills that offer access to work,
the ability to benefit from lifelong and work-based learning, and to switch between a range of
work functions and careers. Several of the guidelines contained within the improving
employability pillar are consistent with this requirement. However, a larger supply of skills
will not raise productivity, growth and employment in the absence of flexible working
arrangements and greater product and capital market flexibility — which is why the guidelines
on developing entrepreneurship and encouraging adaptability are equally as important.
A key aim should be to foster a United States-style entrepreneurial culture. It is thus
encouraging that the developing entrepreneurship guidelines encourage Member States to
improve the availability of venture capital, to reduce overhead costs and administrative
burdens on business, and to reduce tax obstacles to small business formation.
At present, budding entrepreneurs in the United States find it easier to obtain venture capital
than their EU counterparts and face a regulatory environment which makes it relatively easy
to set-up and operate (for example, in terms of allowable hours of operation and planning
restrictions, etc.). This helps increase the contestability and competitiveness of United States
product markets and is arguably the main reason why employment rates in Europe are lower
in all parts of the service sector, high-skilled and low-skilled (27).
This fact should not be overlooked as it is often implied that the United States outperforms the
EU only because it has a very de-regulated labour market and minimal welfare provision that
encourages the creation of poorly paid service jobs. United States’ employers face few
restrictions over hiring and firing staff and can tap into a ready supply of low wage labour. The
minimum wage and labour taxes are low. With limited benefits jobless people have little
option but to take low-paid jobs (many then receiving annual tax credits to boost their
incomes). Consequently, businesses are able to operate and create jobs in low value added
service sectors where profit margins are tight.
However, whilst this undoubtedly leads to the creation of a lot of low productivity jobs it is
necessary to remember that the United States has a relatively large pool of low-skilled labour.
It is arguable therefore that United States-style business practices would be more beneficial to
(26) Seltzer, K. and Bentley, T. (1999). Also Westwood, A. (1999).
(27) European Commission (1998).
282
Weightlessness and the political economy of service sector job creation
the creation of higher productivity jobs in the EU because the EU’s generally superior
education and training systems provide employers will a better supply of skilled labour.
Although a relatively high proportion of people complete upper secondary schooling in the
United States, a relatively high proportion also leave school with poor basic literacy and
numeracy skills (28).
As for the EU’s own pool of low-skilled labour the ‘low pay or no pay’ trade-off can be
alleviated by measures that offer greater incentives to businesses to create job vacancies, and
to individuals to fill them, at existing real wage levels. This objective is pursued by the
improving employability guidelines on activation, and by the developing entrepreneurship
guidelines relating to the possibility of lower labour taxes for low-wage labour and reductions
in VAT on labour intensive services.
1.3.2. The relative merits of private sector, public sector and social economy jobs
Policy-makers may be concerned not just with the overall level of service sector job creation
but also the sectoral distribution, especially in the face of market failure. For example, private
entrepreneurs may fail to supply certain types of social service, be reluctant to set-up in high
unemployment areas and/or be reluctant to employ certain groups, such as individuals with
multiple disadvantages in the labour market.
As a result, it may be deemed necessary, or considered preferable, to create service sector jobs
outside the private sector. The public service sector has traditionally been the main source of
such non-market employment. However, the so-called ‘social economy’ — consisting of nonprofit distributing organisations producing goods and services to meet particular social needs
— is increasingly identified as a potential source of service sector job creation (as evidenced
by the 1999 employment guidelines).
However, policy choices may be constrained. There may, for example, be structural or
financial limits on job creation outside the private sector. Assuming that EU Member States
remain committed to fiscal consolidation within the constraints set by EMU and/or
governments are unable or unwilling for political reasons to finance employment creation by
raising taxes, there is likely to be an imperative for job creation measures that involve low,
zero (or perhaps even negative) net cost in terms of public finances.
In view of this, it is useful to establish criteria against which to assess the contribution that an
expansion of employment in different sectors of the economy might make to overall economic
and social well-being. These criteria considered below are: (1) targeting of jobs, (2) the cost
of job creation to the public finances, (3) productivity, (4) pay and conditions of employees
and (5) scope for job creation.
(28) The United Kingdom Department for Education and Employment and Cabinet Office (1996).
283
The job creation potential of the service sector in Europe
1.3.3. Targeting of service sector jobs
When it comes to targeting, the public and social economy sectors score more highly than the
private sector. Social economy organisations in particular target their services at
disadvantaged groups and distressed communities, often providing jobs to unemployed
people from those communities. Indeed, although cooperatives and mutuals have a long
tradition, much activity in the social economy over the past two decades has been a direct
response by voluntary associations to the unemployment and social distress caused by the
withdrawal of private investment from many local areas, especially where cash-starved public
agencies have struggled to cope with the consequences.
In this latter respect community enterprises have played a particularly important role in
generating local multiplier effects, enabling income to circulate within depressed local
economies that would otherwise have deteriorated still further owing to an ongoing dearth of
private and public investment.
However, whilst generalising about the social economy is fraught with difficulty, it is probably
fair to conclude that social economy organisations target more closely in relation to the areas
they serve and the services they provide than in terms of the jobs they create. For example, a
recent large study of 800 social economy organisations in lowland Scotland — one of the
United Kingdom’s most vibrant areas of social economy activity, employing 42 000 people,
on a par with the region’s electronics sector — found that 79 % of organisations spent at least
half their income on helping disadvantaged groups. However, while 30 % of organisations
recruited at least half their paid employees from disadvantaged areas only 16 % recruited at
least half from disadvantaged groups (29).
1.3.4. The cost of job creation
Private sector jobs are ostensibly created at no cost to the public purse, albeit in the case of
low productivity jobs the State may have to finance supplements to the incomes of low wage
workers. Cash strapped governments are therefore likely to look favourably on private sector
jobs. Similarly, while many social economy organisations are to some degree dependent on
public funds from central and local government and the EU structural budgets, some are selffinanced and others financed from multiple sources, including trading surpluses. In public
policy terms this means that on the face of things, creating jobs in the social economy sector
is also a less expensive option than standard public sector job creation.
1.3.5. Productivity, pay and conditions
Private, public and social economy sector jobs range from high to low productivity activities.
For example, in most EU economies education and health services are located in the public
(29) McGregor, A. (1997).
284
Weightlessness and the political economy of service sector job creation
sector and comprise everything from very knowledge-intensive work to ancillary work. The
same is true of the social economy. The principal areas of activity of social economy
organisations (especially non-profit associations) are social care, education, health care,
housing, culture and recreation. Some of these activities are themselves knowledge-intensive
and even where they are not, social economy organisations tend to be led by skilled social
entrepreneurs.
Nonetheless, at the low skill end of the job spectrum the public and social economy sectors
tend to be more labour intensive than the private sector and thus score poorly on conventional
measures of labour productivity. However, such conventional measures (namely, output per
worker employed or output per hours worked) do not capture the ‘social productivity’ inherent
in the public and social economy sectors. Where the provision of a social service is important,
efficiency and effectiveness (at least up to a point) is raised, not lowered, by increasing the
ratio of labour inputs to service outputs. Mention of social productivity also highlights the fact
that the social economy employs workers with considerable ‘people skill’ of a kind not always
recognised by formal definitions of skills and qualifications (30).
Pay in the private sector varies considerably in line with differences in productivity. However,
it is evident that in both the social economy and public sectors social productivity is not
adequately reflected in pay rates. Even skilled workers in these sectors tend to earn less than
their private sector counterparts. Much instead depends upon the value society places on work
undertaken in these sectors and the willingness of taxpayers to fund pay awards. The social
economy is a particularly low paid sector. In the United Kingdom, the National Council for
Voluntary Organisations (an umbrella group for mostly non-profit organisations) reports that
pay for employees in the voluntary sector is between 20 and 30 % below that in the
commercial sector and usually lower than pay rates in local government and the civil service.
Some social economy organisations deliberately keep pay scales low for ethical reasons in
order not to create a ‘social gap’ between the incomes of staff and those of clients in the
communities they serve. Indeed, some employees in the social economy may even be prepared
to work for a relatively low wage because of a ‘vocation’ to the work they are doing (or
because they derive a high ‘psychic wage’ outside the commercial ‘rat race’ that compensates
for relatively low financial reward). However, while this may be considered laudable from the
point of view of social solidarity there is clearly a risk that organisations might take advantage
of the goodwill of employees, especially given the relative absence of organised collective
labour practices within the social economy. This is particularly true when one considers that
lack of funding creates pressures on social economy organisations to keep core costs low.
(30) Delors, J. (1996).
285
The job creation potential of the service sector in Europe
1.3.6. The scope for job creation in the private, public and social economy sectors
The scope for private service sector job creation is likely to be unlimited so long as the general
conditions for growth and employment in the economy are met. By contrast, funding
constraints are likely to limit the ability of governments to create jobs in the public services.
This leaves the social economy as the main alternative means of service sector job creation.
Therefore, what contribution might the sector make to employment growth?
The social economy has already become a substantial generator of jobs throughout the EU.
The European Commission estimates that in the early to mid-1990s some 6.4 million jobs
were situated in the social economy — 4.4 % of total employment in EU-15 (31). This amounts
to around 10 % of total service sector employment. Given the extent of unmet social needs
and cultural wants that might be catered for by enterprises operating within the social
economy there would appear to be scope to increase employment in the sector. However, it is
necessary to caution against over-optimism.
The European Commission notes that between 1988 and 1992 the annual rate of growth of
employment recorded by the EU’s 500 ‘most dynamic’ social economy organisations was
6.6 % (32). Applying this rate of growth to the current EU stock of social economy jobs
indicates that it would take 12 years to double the existing stock, even if such a rate of
expansion were possible on a sustained basis. This offers some indication of the likely limits
on the overall contribution of the social economy to service sector job creation.
In this respect, however, it is worth noting that the activities identified by the European
Commission as potential sources of social economy jobs range from childcare and domestic
services to housing improvements, cultural development and nature protection. It is not
immediately obvious that activities of these kinds are sector specific. It might be possible
therefore to devise innovative partnerships between private, public and social economy
sectors to create such jobs, thereby maximising the job creation potential of limited public
funds. Indeed, insofar as joint ventures of this kind might themselves be facilitated by
information networks, it is possible that knowledge technology will prove its ability to help
meet social need as well as being a source of overall growth and employment in the new
weightless economy.
(31) European Commission (1997).
(32) Idem.
286
Chapter 2
Beyond conventional service economics —
utility services, service product chains and job
services
Gunther Schmid
2.1. Introduction
Two pathways to increase service employment are usually suggested: improving
competitiveness of high quality and expert-oriented business services, and relieving the cost
burden for (low-skill) personal services. Whereas competitiveness of business services can be
increased through innovations in information and communication technologies (innovative
milieus or networks), low-skill personal services are thought to be stimulated by widening the
wage spread combined with various wage subsidies or income policies. There are still
untapped possibilities for improving these strategies by policy reforms or further research.
However, both pathways are rather conventional and do not exhaust the range of innovative
approaches towards the modern service society.
In the following essay, I will therefore explore three complementary (if not alternative) ways
to foster job creation in services: First, the switch from product markets to utility markets
would induce more and new types of user services. Such a switch would not only be more
employment-friendly but also foster a sustainable economic development compatible with
ecological principles. The reorientation from products to utilities provides strong arguments
for an emphasis on the framework conditions of services; for instance for lowering VAT in
favour of labour intensive services, for forward regulation through (usually service
enhancing) high quality standards, for deregulating market entry barriers, and for subsidising
networks instead of individual products or services. Second, Baumol’s cost disease has to be
reconsidered in view of productivity-enhancing possibilities by new technologies, and in view
of innovative possibilities in managing specific income risks related to modern services.
Service product chains provide the main ingredient for increasing productivity in services, and
I will demonstrate this in the field of art services, Baumol’s prototype and the inspiring area
287
The job creation potential of the service sector in Europe
for the theory of cost disease. The evolution of social insurance in this field also offers
intriguing ideas for risk management beyond the traditional welfare state. The most important
policy conclusion related to this experience is the inclusion of consumers into the financing of
social insurance. Third, the switch from transfers to persons with low earning capacities to
supporting people’s self-reliance through their own work could also induce increasing demand
for already existing or new types of job services, and contribute especially to gender equality
in the labour market. The central policy conclusion is the recommendation to turn social
benefit entitlements (transfers) into vouchers for buying or co-financing job services.
2.2. From producer services to user services
So far, high quality services have been looked at mainly from the perspective of productivity
enhancing inputs into (manufactured) products. However, this perspective becomes more and
more elusive, as the rationalisation of services themselves is only to a limited extent possible.
Thus, if the input of services for a product (research, development, raw material transport,
labour services, marketing, advertisement, transport, distribution and storing of the ready
product) and finally the required waste disposal of the used product reaches a certain
proportion, the price of the final product might become so high that it is not marketable, even
if the production process itself is completely automated and at the highest productivity level.
When the costs of all services necessary for distributing and using the product are increasing
much faster than the increase of productivity on the pure production side, the overall
productivity curve points downwards before reaching the top of the S-curve in the traditional
theory of ‘economy of scale’ (Figure 2.1).
Figure 2.1: The productivity paradox
(a) Increases in productivity regarding the effects of new technologies and specialisation
only,
(b) and, in addition, taking into account all service functions necessary to make the product
available to the customer
Pi: Productivity increase
Source: Adapted from Giarini and Stahel, 1989, p. 18.
288
Beyond conventional service economics — utility services, service product chains and job
services
A model case is the idea of disposable sheets (reported in Giarini and Stahel, 1989: 18). On
average, a sheet is washed about fifty times a year before being thrown away. For producers,
it would be fantastic if consumers would buy fifty sheets instead of one, attracted by the much
lower price and by the elimination of the necessity of washing them. However, this would
mean storing an enormous number of sheets at each home, the destruction of the same amount
of sheets either at home (at a cost very near that of washing them), or through an addition to
the local waste disposal systems (which would increase local taxes for collecting and
destroying the bedsheets). The lesson of this model is that the key economic problem is to
consider the optimum product-life of a specific good, taking into consideration the costs of all
services related to utilise and finally to dispose of the consumed product. The utilisation or
function of the product is the real issue, and products have to service optimal utilisation, and
not the other way round, for services to increase the amount of products.
How such a shift in the production system towards a real service economy would affect
employment in favour of both low-skill and high-skill services (and in the regions where
people are living) can be demonstrated by the case of extending the product-life of
automobiles (Figure 2). In the specific example, the pattern of which can certainly be
generalised, the purchasing price of the car after 10 years of use makes up 58 % of the total
costs for the consumer; the labour costs concerning maintenance and repair are at 20 %
relatively modest, and so is the 17 % expenditure for replacement units. After 20 years of use,
however, the (same) purchasing price’s share of the total costs declined to 32 %, whereas the
share of labour costs increased to 33 %, and the share of replacement units to 29 %. After 20
years, the utilisation of the car contains higher expenditures and therefore higher investments
in the form of local wage work than in the form of manufacturing the final product. In
addition, the costs for replacement units and consumption material together exceed the
purchasing price.
Increasing the life of products by increasing their time of utilisation, therefore, increases the
share of service employment in the local region of utilisation. In addition, increasing productlife substitutes energy and raw material with labour (preserving, thereby, the natural
environment), whereas replacement units will come partially from dismantled old cars and
partially from suppliers who are independent from the automobile producer. The producers of
new cars and people employed by them are the losers in this game since the turnover of
automobile products and corresponding sales fall. It is very likely, however, that the net
employment effect for the overall economy is higher since the new services are more labour
intensive than the car manufacturing firms which might, in the extreme case, be totally
automated.
289
The job creation potential of the service sector in Europe
Figure 2.2. Comparison of energy and labour input for manufacturing and maintenance of a middle
class car (Toyota Corina MK II, 1969) with a utilisation period of 10 and 20 years; without gasoline
Source: adapted with small corrections from Stahel, 1997, p. 87.
For the user, the longer utilisation of the car must not be a loss of utility, if the design of the
automobiles is such that one can keep track of the latest useful technology by replacing
corresponding components. One can even imagine that only new technology which really
improves utility will be stimulated by such customers who — not having to buy new cars —
can afford more new components. A further condition, of course, is that the prices of
maintenance and repair services are affordable. This, probably, requires new incentive
structures by the tax system, for instance a differentiation of VAT (value added taxes) and/or
a differentiation of employers’ contribution to the social security system (the so-called nonwage labour costs) in favour of services. From this point of view, the proposal by the European
Commission to differentiate VAT in favour of service employment makes much sense
(Europäische Kommission, 1999). In addition, the regulation of high quality standards aimed
at extending the product-life would enhance high quality services or the content of services in
the products. A case of such ‘forward regulation’ would be the obligation of car producers to
take back cars after utilisation. Finally, more and more specialised services are to be expected
290
Beyond conventional service economics — utility services, service product chains and job
services
which is a case for deregulating any market entry barrier through occupational membership
rules, and replacing such rules by professional standards supervised by a licence system.
Such a new strategy would replace consumers by prosumers (Alvin Toffler) who buy utility
systems and not individual goods (for example, leasing transport facilities instead of buying
cars which are not used 95 % of the time), and it would replace producers of goods by the
managers of utility systems. The new service economy is a complex production system in
which supply and demand are not strictly separable as traditional economics supposes; in such
a system, the distinction between manufacturing and services becomes more and more
blurred (1). The new service economy aims at providing results or an economic value defined
in terms of actual utilisation. From the ‘production side’, this includes various inputs
(maintenance, counselling, repairing, restructuring) during the time of utilisation plus finally
the prospective costs of waste management related to the product. From the ‘consumption
side’ this involves contributions by the consumers themselves as part of the production
system. Product managers inevitably encounter the situation in which they need the active
collaboration of the users who condition the possibilities of generating usable results. The user
or consumer thus becomes her/himself a condition for making the ‘production’ work (Giarini
and Stahel 1989:94).
As the contribution of the prosumer is usually an activity or work which cannot be
monetarised, the distinction between paid work (Lohnarbeit or labour) and unpaid work
(Aktivität or work in the narrow sense) also gets blurred. In any case, the new service economy
requires the explicit articulation of the interface between the labour market and the nonmonetarised world of work. Paid work (services) for childcare, for instance, will have to
compete with unpaid (services) for childcare, and the choice between both types of ‘work’ is
not based on comparing prices with shadow prices but is according to a more complex
procedure of evaluation. This procedure includes indicators of performance or quality in terms
of educational performance and sociability as well as valuing the emotional rewards. The
outcome of such procedures, as we know, might be a combination of both ways of providing
childcare depending on culture and budget constraints. The value of health services, to give
another example, cannot be determined by the consumption of drugs or by the salaries of
physicians, but in terms of the level of health achieved for a given population or individual;
the same holds true for educational services. Whereas during the industrial period the value of
products could be identified essentially with the costs involved in producing them (material,
wages, capital), the notion of value in the service economy is shifting towards the evaluation
of costs incurred with reference to obtained results in utilisation (Giarini and Stahel,
1989: 33).
In the extreme case, the price of the product might even become zero or a symbolic sum like
the case at present in Germany in which high quality computers are offered for only one DEM,
however related with buying a whole package of services for the use of Internet or specific
(1) This is also one of the main conclusions from many sophisticated comparative empirical studies, for instance
by Peter W. Daniels (1999).
291
The job creation potential of the service sector in Europe
software. In Sweden, mobile phones may cost only a crown but are coupled to a telephone
subscription. In a way, one can interpret this product-service chain as a form of credit market
in which one receives the product as credit and pays later for the related services (or immediate
utilities) (2). Where high capital costs are involved, leasing is one appropriate form of buying
utilities, and product-sharing (like car-sharing or vacation-home sharing) another. In both
cases, various services are involved which are not required in a simple product-consumer
chain: organisation, maintenance, a more complex form of financing and insurance, contract
management. Functioning product-service chains have job creation potential, because
products are to some extent substituted by services which — normally — are more labour
intensive than the products.
Thus, the term ‘services’ means, finally, performance in real periods of times whose value over
the time is uncertain and has to be determined by probabilities. If this is accepted, a theoretical
framework of reference based on the notion of risk and uncertainty is required, and policy
interventions or institutions suitably adjusted to the new service economy will have to
consider the insights of modern risk management or insurance policy. If not, products but
service packages or utilities over time are purchased, it is clear that conflicts arise about the
right value and the need to insure against the risks that the service is not delivered as expected.
As the new service economy also implies increasing horizontal complexity (interaction
networks), service production and delivery systems become more vulnerable to small errors,
unknown influences from outside or even criminal attacks. This explains the rocketing rise of
various insurance and counselling services. As we know from history, innovation in risk
management was always extremely important for economic development (3).
To sum up, increasing the job creation potential of services requires, first of all, an emphasis
on changing the framework conditions in favour of immediate utilities (functions or results)
instead of products with potential utilities. Such a switch increases the consumer’s sovereignty
but also the consumer’s responsibility in participating in the valuing process; very often it is
only this interaction between the consumer and the producer which brings forward the
decisive value added of the service involved. Behavioural customs or transaction costs, of
course, will slow or — in some areas — even prevent such a transition from consumers to
‘prosumers’. However, as there is no ‘free lunch’, there is also no ‘free utility’. Policy can only
set the right incentive structures. However, there are — as I have tried to demonstrate — still
many untapped possibilities in this respect.
So far, I have argued on the basis of Baumol’s cost disease which means that important
productivity gains are not possible for most of the labour intensive services. Thus, if the
content of such services in products increases, the price of the product may become
prohibitive and therefore not marketable. The switch to utility systems was a strategy to
overcome such a stalemate. However, at a closer look, more services than conventional
(2) I owe this idea to Dominique Anxo and Donald Storrie.
(3) For an enlightening and extremely well written story about risk management in history, see Bernstein 1996.
292
Beyond conventional service economics — utility services, service product chains and job
services
economics imagines are prone or suitable to productivity gains. As shown in another
contribution (D. Dathe and G. Schmid, 2000), interaction between knowledge-intensive
manufacturing and knowledge-intensive services (‘service-industry districts’) explains much
of the divergence in regional employment performance. In the following, I will reconsider
Baumol’s cost disease in his own favoured field, in the area of art services, a case however,
which can be generalised beyond that (4).
2.3. Repealing cost disease through service product chains
For Baumol, productivity gains have created ways to reduce the labour needed to produce a
car, but no one has yet devised ways of reducing the work hours needed to perform a 45 minute
Schubert quartet (Baumol and Bowen, 1993 (1966): 164). Baumol’s ‘sombre conclusions’ are
clear: The arts cannot match the productivity gains of the economy as a whole. As the
musicians performing this nice quartet or other pieces ask for wage increases related to the
overall productivity, a chronic gap between the costs of performing organisations and their
income is created. Either art disappears slowly or subsidies, tax exemptions or other incentives
have to fill the productivity gap. Apart from this cost disease, Baumol cites the social benefits
(especially in form of positive externalities) as argument for subsidising the arts. Since
education, child or health care are also prone to the same cost disease, large areas of human
services require continuous public support in one way or the other.
The logic of this argument is unassailable, and it certainly still holds true for a substantial part
of human services. However, the assumptions need both empirical and theoretical reexamination (5). David Throsby (1994) reviewed empirical studies finding lower wage
increases (than overall productivity increases), changes in the repertoire (which requires fewer
players or low copyright fees), changes in the organisation of performing (larger concert
halls), increasing demand (income elasticity in favour of services), and — finally — donations
as mechanisms for increasing productivity or overcoming the cost disease. More importantly,
technology, specifically the development of mass media, rendered the private good of
orchestra music into a potential public good which is unrivalled in consumption at various
levels: as live performances, as broadcasted performances, and as recorded performances.
When a 45 minute Schubert quartet is played, (almost) no additional production cost is
incurred by additional listeners (Tiongson 1997: 119). Although Baumol and Bowen already
acknowledged these possibilities in their classic book on the economics of art, they
underestimated their impact.
Apart from the aforementioned possibilities of productivity increases, the cost disease can be
overcome specifically through service product chains in which the one-time performance
(4) The following section has been inspired by Frank and Cook 1995, Menger and Ginsburgh, 1996, and especially
by Tiongson, 1997 who reviewed some of the recent literature on Baumol’s cost disease in the economics of
art. For the labour market of artists and journalists in Germany see Haak and Schmid, 1999; for an excellent
monograph on the economics of culture (with special emphasis on France) see Benhamou, 1996.
(5) For a systematic critique of the Baumol model see Tronti, Sestini and Toma 2000.
293
The job creation potential of the service sector in Europe
induces a chain of other services or products which again feed back to raise reputation and
demand. Broadcasting and recording may function like advertising, making it more
prestigious to consume the orchestra’s products in the process. In addition, merchandising art
activities may provide substantial additional income, for example, from Broadway — shirts,
posters, and other souvenirs. Christo’s art of packaging is a clear model for such service
product chains where the manufactured products at the end pay for the performance. Because
merchandising activities benefit from advancements in the manufacturing sector, this suggests
other, more complex relationships between the arts and the rest of the economy. The economy
uses the arts more and more, as the arts are using the economy. Technology may facilitate
productivity gains in both the manufacturing sector and the performing arts, albeit in different,
and at times interconnected, ways (Tiongson, 1997: 120). Symphony orchestras or theatres
can develop into multi-product non-profit enterprises with performances varying in content,
audience, composition, location, and number of performances; or they can develop ‘strategic
alliances’ with commercial enterprises.
The logic of service product chains is not only restricted to the sector of arts. The digital
revolution opens many possibilities to transform originally expensive services into
standardised cheap products later. The sector of education is a good example in which learning
programmes (software) replace the immediate act of teaching, or at least part of this face-toface service. The genuine service act of teaching can be specialised and be relieved from
routine tasks. Like the product-services chain discussed in the first section, the service product
chain can open a symbiotic relationship with overall higher productivity plus increased
employment if new markets are opened through this (for instance professional training
services in computer aided design, training in multimedia presentation). Public provision or
publicly supported provision of the required communication networks (information highways,
computing facilities in schools, etc.) are crucial to develop such potential new markets.
Although service product chains are flourishing and still open untapped possibilities, the fact
cannot be contested that the majority of artists or artist-like professions face lower income or
at least much more volatile income streams than the average wage earners of comparable
(formal) skill levels (6). This is largely due to the ‘winner-take-all’ quality of these markets,
which has been studied perceptively by Frank and Cook (1995). Menger and Gurgand (1996)
provide additional arguments for the structural surplus of labour in the artist labour markets,
which are responsible for low and volatile wages for a majority of artists. Haak and Schmid
(1999) show the dual structure of income distribution among self-employed artists in
Germany — a relatively high percentage in the bottom income range, and a relatively high
percentage in the upper range, but a flat curve around median income levels instead of a
‘normal distribution’. The income structure of dependent artists does not differ much from that
of the total active labour force, however, considering the above-average education level of
artists, one would expect a higher average income (Figure 2.3).
(6) For the two different but functionally equivalent models of artist labour markets in the United Kingdom and
France see Benhamou and Gazier, 2000.
294
Beyond conventional service economics — utility services, service product chains and job
services
Figure 2.3. Monthly net income of artists compared to all employees
Source: Micro Census 1995; own calculations (Haak and Schmid, 1999).
How do (self-employed) artists solve the problem of high income risks involved in their kind
of service activities? Are there comparable patterns in other professions? Can we learn from
these institutions of risk management if those kinds of service jobs are further increasing or
even intended to be fostered? An interesting model is the French special unemployment
insurance system for artists as described and explained in Menger and Gurgand (1996). This
system provides relatively generous benefits during the intermittent periods of nonengagements, a system heavily subsidised by the State. Since its introduction during the
1970s, the number of artists as well as the number of (short) engagements increased due to
serious moral hazard problems. Although one could legitimise the subsidies, along the line of
Baumol, with the argument of cost disease and with positive social externalities related to an
increasing number of artists, it seems unlikely that such a system can survive (7).
The German social security system for artists and journalists, introduced in 1982 after long
and heavy battles (8), is far from perfect but provides an interesting innovative feature which
could serve as a model for larger welfare reforms. Self-employed artists are dealt with like
dependent employees in as far they were obliged to pay contributions into the ‘art social
security fund’ (Künstlersozialkasse, KSK) which should cover 50 % of the expenditure. The
(7) See also Benhamou and Gazier (2000) for the most recent development of this system.
(8) In which the recent Nobel price winner Günter Grass played an active role in pushing through the new system.
295
The job creation potential of the service sector in Europe
innovation is, that the other 25 % are subsidised by the State (9), whereas 25 % comes from a
special tax of art utility providers, merchants or publishers; for instance art galleries have to
pay 4 % of the honorariums or the sales to the KSK. The KSK covers health risks and pensions
but not unemployment. It is a true public-private mix, and the implicit participation of the (art)
consumer takes into account the asymmetry of ‘winner-take-all markets’. As more and more
jobs become like the jobs of artists or journalists (Haak and Schmid, 1999), an extension and
continuous reform of this system seems to be recommendable.
There is, finally, another aspect of risk management in the artist labour market which is closely
related to the uncertainty of the utility value which we considered in the first paragraph. Artist
or writers in the service product chain produce prototypes whose utility for consumers is (by
definition of the art as an institution specialised in innovation) not known and valued
consecutively by market processes. The common private solution to this uncertainty is
authors’ property rights and royalties that guarantee income or profit-sharing depending on the
‘result’ of the valuing process (10). How effective this ‘future market’ is working seems not to
be well studied so far. It seems plausible, however, that equity and fairness require regulatory
provisions that compensate for the often structurally weak position of individual artists or
writers or professions in a comparable situation. Successful artists are able to work with and
to pay for professional agents, a precondition which many young artists, beginners or not yet
successful artists are not fulfilling. Trade unions may find here an untapped area for collective
support and specialised advice or services. However, as artists provide services or produce
products which often have the character of public goods, apart from the engagement in social
security (see above), public provision of infrastructure and (regulated) obligatory risk sharing
systems seem to be required to mobilise the job creation potential in artist and artist-like
labour or product markets (11).
3.4. From personal transfers to job services
As we know from modern job search theory, social networks play a crucial role in adjusting
to structural changes in the labour market, especially in finding a new job (Anxo and Storrie,
1999). The policy consequences of this insight have not yet been rigorously addressed.
Although generous cash benefits can be regarded as financial precondition to maintain
networks (membership in clubs, attending social events with friends etc.), a more proactive
orientation towards establishing or supporting existing networks can be taken. It is almost by
(9) The present government intends to reduce this share to 20 % due to heavy deficits of the public budget, but also
to prevent further increases of members in the art social security fund which, by the way, holds also for
journalists or writers.
(10) See on this point also Benhamou and Gazier, 2000.
(11) Here is a structural equivalent problem to ponder: Why should firms or individuals invest in the production of
knowledge (or art) if they cannot exclude others from the final product? The American economists Romer,
Hyde and Solow, basing their argument on experiences in Silicon Valley, argue for deregulation in the sense of
legally not supporting property right issues related to knowledge, and for generous public support in various
ways, among others fostering private art sponsoring through tax exemptions or — this provides a bridge to the
next part — by extending the earned income tax system (Hyde, 1998).
296
Beyond conventional service economics — utility services, service product chains and job
services
definition true that ‘social networks’ cannot be created artificially. ‘job search courses’ and the
related job services, therefore, are often rightly ridiculed. However, if one regards ‘social
networks’ as a metaphor for the existence of ‘experience knowledge’ and ‘access of relevant
information,’ there are several ways to exploit this insight in a more systematic way. Some of
this ‘experience knowledge’, often not known by the local placement officers, can be
stimulated by arranging suitable context conditions which usually involve professional
services related to job search. The activating of cash benefits into job (search) services makes
also sense from the knowledge that service jobs, especially new services related to the
information sector, are often ill-defined and changing fast. Thus, the importance of experience
knowledge is itself related to the service economy and its development.
In the following, I will briefly sketch some examples and good practices related to the basic
concept of stimulating job services as a kind of ‘double strategy’: to increase employment in
a relevant and increasing segment of the labour market which enhances the efficient job search
or job mobility itself. I will thereby apply a very wide (and contestable) definition of ‘job
services’ which ranges from ‘self-servicing’ through accepting low-wage jobs, up to the top
with a kind of negative income tax to job services provided by commercial time work
agencies.
(1) As most jobs are related to a ‘social network’ (co-workers, clients, a new neighbourhood),
widening the range of acceptable jobs can be counted as ‘job service’, because accepting
a job sets in motion a ‘self-servicing’ activity to collect relevant experience knowledge.
The acceptance can be blocked by high reservation wages, lack of job-related skills or the
loss of cash benefits. In any case, the possibility of transforming the cash benefit into a
voucher to be used as wage (cost) subsidy or, if such an entitlement is not at stake, by
compensating the deficits of earning capacity by a kind of negative income tax (like the
Earned Income Tax in the United States would increase the range of choices of individuals
without a job and certainly increase their self-reliance.
The institutional details are important but cannot be discussed here in detail. In general,
however, it is obvious that such a scheme cannot be universal to avoid market distortions.
Compensation of earning capacities seems only legitimate if connected with socially
acknowledged (legitimate in terms of values) reasons. Candidates for such targeted income
compensation schemes are people with care responsibilities for dependants (relevant
especially for men and women with small children) or persons with work disabilities. In some
cases, this kind of ‘job service’ might involve the complementary creation of ‘public jobs’ to
provide the opportunity of choices. However, because such jobs often provide a narrow or
stagnant ‘social network’, such a ‘secondary labour market’ should be kept very restricted.
Another, more relevant case, is the provision of public infrastructure (daycare for children
etc.). Such a system will not come cheap. However, as Bob Solow (1998) persuasively
elaborated, each person taken from the welfare payroll increases altruism at the same time,
which means the willingness to pay for such a system. Self-reliance and altruism are mutually
reinforcing.
297
The job creation potential of the service sector in Europe
(2) Placement services by the public employment service (PES) and or by private
employment services (PRES) are acknowledged in the evaluation literature to be the most
cost-efficient measure of active labour market policy (Walwei 1996). Training measures
or wage subsidies turned out to be only effective under specific circumstances, very often
connected with an effective placement service before or after the measure. Thus, to some
extent a shift from cash benefit expenditure or from less effective job services to
placement services seems to be justified. At present, only between 10 and 30 % of the total
hirings in Europe are supported through intervention by PES. In 1995, around ECU 185
billion (3.3 % of Union GDP) was spent on income support for the unemployed and active
labour market programmes; ranging from counselling and brokerage to training and
subsidised employment. One third of this sum (ECU 65 billion) went for active measures,
which to a large degree was organised through PESs. The share of national GDP spent on
active labour market programmes ranged from 0.3 % in Austria to 3 % in Sweden. Across
the Member States expenditure on PESs’ staff and equipment ranges from less than
0.10 % to around 0.25 % of GDP. This range reflects individual Member States’
expenditure on active and passive labour market programmes and the part played by PESs
in implementing these programmes. Excluding the administration of benefits and
programmes, expenditure on PESs is estimated to be below 5 % of total expenditure on
active and passive policies. This is a fairly modest investment which can bring a
substantial return in terms of increased market efficiency and of savings on
unemployment benefits (European Commission 1998). An increase of this kind of job
service — under the condition of a modernised PES (through professionalisation,
decentralisation and cooperation with private professionals) — would certainly pay off.
For Germany, it had been calculated that reducing the average duration of unemployment
(presently at a level of almost eight months) by one month through intensification of
placement services would save around DEM 11 billion on cash benefits and related
transfers (Schmid et al., 2000) (12). Research on determinants of mismatch shows that it is
especially services which are prone to long unfilled vacancies.
(3) There is an untapped potential for initiatives based within local communities which may
aid in the formation of networks. The local communities would also probably be more
capable than the PES to mobilise and to interact with non-governmental organisations
(NGOs). City employment contracts, and local or regional employment pacts for
organising local job markets have turned out to establish quite efficient networks for
various employment initiatives. It is self-explanatory, that the implementation of such
networks is job-intensive, and therefore not a cheap solution. However, since functioning
networks have long-term impacts, such investments usually pay off (13).
(12) It would, of course, be nice to have harder facts such as some kind of ‘elasticity measure’ showing how much
expenditure on job services for placement activities is necessary to reduce cash benefits for the unemployed
by 1 %. I do not know such figures or estimates, therefore my argument is (I guess, however, reasonable)
speculative and suggestive.
(13) For the overall philosophy of the ‘network society’ see Castells, 1996; for theoretical backgrounds and
examples of regional employment policy see Sabel, 1989; for the theory and practice of policy networks see
Marin and Mayntz, 1991; the importance of the network strategy for rural areas is demonstrated in Van
Depoele, 2000.
298
Beyond conventional service economics — utility services, service product chains and job
services
(4) Buying professional placement and training services is another untapped source of
transferring cash benefits into job services. One strategy is the internalisation of
placement services into firms and using benefit entitlements of the potential unemployed
as co-financing measures for re-training or wage subsidies related to a new job. For
instance, outplacement services in cases of large-scale restructuring or bankruptcy and job
rotation are used more and more as efficient instruments, and they slowly create a market
for commercial services in this field. Another strategy is bundling interrelated services to
tap spillovers or economies of scope: PES contracting with professional private training
centres can include placement into jobs as an explicit element of the service package and
thus stimulate more market-oriented training courses. The establishment and support of
training and education networks between firms which alone do not have the capacities
would also be a new form of job services. Availability of local job services is important
especially for young couples tied to family responsibilities at home. Belonging to the
same category is the purchasing of rehabilitation services and job maintenance services
for injured persons, chronically ill persons, and elderly people with reduced work
capacities. Also job-pools and the cooperation with temporary work agencies are
promising job services which can substitute cash benefits. Experience in Germany and the
Netherlands shows that between one and two-thirds of temporary workers find a regular
job this way (14).
(5) A final case — closely linked with the aspect of uncertainty related to service utilities
considered in the second section — is the support of risk management services related to
job creation. An example is the establishment of networks and specialised business
services supporting business start-ups for the unemployed. In comparative research, it has
been found that the evaluation of business start-ups for mobilising the required capital is
much more future- (or forward-) oriented towards the potential utility of the new business
idea in the US than in Germany (15). In the German system, the evaluation of business
start-up ideas is much more backward-oriented in emphasising the financial securities
which the persons or the group applying can guarantee. Mobilising venture capital for
business start-ups is usually the task of professional intermediaries in the United States
whereas in Germany, banks or saving institutions (Sparkassen) are taking over this
business. Banks apply conventional financing techniques relying on traditional securities
(real estate, capital assets). Intermediaries, for instance community development
corporations (CDC) rely on the evaluation of the business idea and the underlying
management concept. New financing concepts such as ‘peer groups’, ‘loan loss reserve
funds’ and ‘stepwise lending’ manage a joint risk sharing between lenders and users. This
demonstrates again the importance of new institutions of risk management which are at
the core of an employment-friendly service economy.
(14) For the Netherlands see Visser and Hemerijck, 1997; for Germany see Rudolph and Schröder, 1997.
(15) See Kritikos and Wießner, 1999 and United States small business administration, 1999.
299
The job creation potential of the service sector in Europe
3.5. Conclusion
The main lesson from this essay starts with the critical difference between manufacturing and
services: Whereas services provide immediate utilities to known clients, manufactured
products provide potential and standardised utilities sold on an anonymous market. If service
utilities are standard and quickly consumed like fast food, window cleaning or transporting
beer, the valuing of these utilities is also standard and most efficiently organised by the
market. The majority of services, however, are utilities which are hard to measure. Their real
value is uncertain, especially if they contain investive elements like health, childcare,
education, information and research services. The utilities of such services are often
accumulated in the future, or are realised only after some years or, in the worst case, never.
Most importantly, many service utilities are often realised only if they are combined with other
services or with non-monetary inputs of the consumer or client (16). So, the problem of these
kinds of services is how to organise a proper valuing system which takes into account the
value of the real outcome or performance which contains high risks and uncertainty, and
which takes also into account the often required non-monetary work to tap the potential utility
of high quality services. Risk management and risk sharing is the real economic problem of
the new service economy. The main thesis of this essay, therefore, is that it is those societies
which display successful employment performance and which have established effective and
efficient risk management systems which are able to properly evaluate contingent and future
oriented utilities.
The chapter considered three strategies which could — over and above the conventional
recommendations — help to enhance a more employment friendly service economy in the
European Member States: first the switch from product markets to utility markets (1), second
the support of service product chains (2), and third the stimulation of job services (3). In the
following conclusions, I concentrate on their specific contribution to risk management and to
their possible impact on employment.
(1) One strategy of enhancing utility markets would be to increase the life cycle of consumer
products such as motor cars, electrical machines in the household and so on. Such a strategy
would, as I have demonstrated with the example of our most celebrated consumer product —
the automobile — affect employment in favour of both low skill and high skill services, and
it would at the same time enhance the regional content of employment and counteract spatial
concentrations of production. I have also developed arguments, that such a strategy would
foster technological development which is more consumer-friendly; this may open new
markets and exportable services for the region. A further condition, of course, is that the prices
of the increased service content of products, like maintenance and repairing services, are
affordable. This probably requires new incentive structures by the tax system, for instance a
differentiation of value added taxes (VAT) or a differentiation of employers contribution to the
(16) To give an anecdotal example: A pain in the hip can be the starting point of expensive treatments by
orthopaedists or surgeons and might never be cured, but in a lucky case, it can be corrected by a chiropractor
and the daily physical exercise of the service consumer, here the ‘prosumer’.
300
Beyond conventional service economics — utility services, service product chains and job
services
social security system (the so-called non-wage labour costs) in favour of services. From this
point of view, the proposal by the European Commission to differentiate VAT in favour of
service employment makes much sense. In addition, the regulation of high quality standards
aimed at extending the product-life would enhance high quality services or the content of
services in the products. A case of such ‘forward regulation’ would be the obligation of car
producers to take back the cars after utilisation. Corresponding regulation might be useful for
the lots of electrical machines in private households. Finally, more and more specialised
services are to be expected which is a case for deregulating any market entry barrier through
occupational membership rules and replacing such rules by professional standards supervised
by a licence system.
Another possibility of enhancing utility markets, not discussed here at length, would be the
switch from input payment systems to outcome payment systems or — at least — to proper
mixes of both. A good example is changes in the payment systems of professors and teachers.
We observe more and more complex evaluation systems which try to assess the real utility
which teachers or professors are delivering, whereas former systems have mainly relied on
input criteria like formal educational status, age, time worked and so on. A further important
element of risk management institutions, also not dealt with explicitly here, are wage
formation institutions. My hypothesis is that payment systems of the new service economy are
outcome or forward oriented, whereas payment systems of industrial societies are basically
input and backward oriented. The failure of Germany to lead the modern service economy has
much to do with its former success in manufacturing and with the still central role of
manufacturing. The input oriented payment systems of manufacturing still govern the service
sector to a large extent where an increasing share of performance oriented payment systems
containing risk sharing elements would be necessary.
(2) The second strategy starts from a critique of Baumol’s ‘cost disease’ by demonstrating that
service product chains are an important way to repeal this ‘iron law’ of services. Here, the
principle of risk sharing is the postponement of the ‘proper’ pay of the service delivered and
the profit sharing from related and succeeding products. The model of the service product
chain demonstrated in Baumol’s favoured sector, the artist labour market, is applicable to
other service sectors, for instance education. The main proper policy response, here again, is
not selective intervention but setting the suitable framework conditions for the functioning of
such networks; for instance by regulating the complex issue of property rights involved in this
issue, legal protection of these rights and the provision of a supportive infrastructure. The
same holds true for the reverse case, the product-service chain in which the basic principle of
risk management is crediting future services related with the (cheap) purchase or leasing of
the product.
To generalise this point, risk management and related jobs will no longer be secondary
services but core activities in the new service economy. Here, neither the smooth market
exchange of individual products nor services is representative, but complex and selforganising provider systems such as ‘health maintenance organisations’ (HMOs in the United
States), ‘human resource management systems’, ‘house energy saving agencies’, ‘local
service agencies’ and so on. Thus, the right policy direction to foster corresponding service
301
The job creation potential of the service sector in Europe
employment should not be to subsidise single services or products but to endorse the
functioning of such complex and self-organising networks (service product or product-service
chains). To give an example: To foster employment in renewable energy systems such as solar
techniques, the proper strategy would not be to subsidise the investment in producing new
solar cells but to support local networks of counselling and risk management in energy saving,
including the cooperation between large-scale producers or utility managers and individual
‘prosumers’.
Apart from these possibilities, the art sector was identified as a prototype of increasing income
risks related to precarious employment relationships. It is likely that such relationships are
spreading more and more to other sectors although there is still need of more precise and
longitudinal analyses in this respect. How do (self-employed) artists solve the problem of high
income risks involved in their kind of service activities? Are there comparable patterns in
other professions? Can we learn from these institutions of risk management if we want to
expand those kinds of service jobs? It was argued that the German case of
Künstlersozialversicherung provides an interesting model of a future oriented ‘employment
insurance’ (replacing old-fashioned UI-systems) for the systematic inclusion of consumers —
the final evaluators of service utilities — in the social security contributions.
(3) The third strategy of switching from cash benefits to in-kind benefits, here job services,
starts from the observation that social networks play a crucial role in adjusting to structural
changes in the labour market, especially in finding a new job. The policy consequences of this
insight have not yet been rigorously taken up. Although generous cash benefits can be
regarded as financial precondition to maintain networks (membership in clubs, going to social
events with friends etc.), a more proactive orientation towards establishing or supporting
existing networks can be taken. Some of this ‘experience knowledge’, often not known by the
local placement officers, can be stimulated by arranging suitable context conditions which
usually involve professional services related to job search. The activation of cash benefits into
job (search) services also makes sense from the knowledge that service jobs, especially in the
new services related to the information sector, are often ill-defined and changing fast. Thus,
the importance of experience knowledge is itself related to the service economy and its
development. Concrete policy proposals reported in this essay were related to targeted
vouchers or earning credits, intensification of placement services, support of local community
networks, buying professional placement or training or rehabilitation services, and the support
of new risk management services related to job creation — for instance business start-ups for
the unemployed.
302
Bibliography
Agell, J., Lindh, T. and Ohlsson, H. (1997), ‘Growth and the public sector: a critical review
essay’, European Journal of Political Economy, 13, pp. 33–52.
Agénor, P.-R. and Aizenman, J. (1997) ‘Technological change, relative wages, and
unemployment’, European Economic Review, Vol. 41, pp. 187–205.
Albach, H. (1989a), Dienstleistungen in der modernen Industriegesellschaft (Services in
modern industrial society), Beck, Munich.
Albach, H. (1989b), ‘Perspektiven des Dienstleistungssektors in der Bundesrepublik
Deutschland’ (‘The future of the service sector in the Federal Republic of Germany’), in
Deutsches Institut für Wirtschaftsforschung (ed.), Dienstleistungen — Neue Chancen für
Wirtschaft und Gesellschaft, 3. Berliner Symposium zur Zukunft der Industriegesellschaft,
Vistas, Berlin, pp. 17–48.
Albrecht, C. and Schmid, G. (1985), ‘Beschäftigungsentwicklung und Qualifikationsstruktur
in Berlin 1977–1983’, Tabellenband (‘Employment dynamics and qualification structure in
Berlin 1977–1983’), Discussion Paper IIM/LMP 85-4, Wissenschaftszentrum Berlin für
Sozialforschung.
Anxo, D. and Storrie, D. (1999), ‘Employability, job search, employer recruitment strategies,
social networks and social exclusion’, in Gazier, B. (ed.), Employability: concepts and
policies, Institute for Applied Socioeconomics, Berlin (Employment Observatory Research
Network Report 1998), pp. 221–243.
303
The job creation potential of the service sector in Europe
Anxo, D., Flood, L. and Rubery, J. (1999), Household income distribution and working time
patterns: an international comparison, Graue Reihe, Des Instituts Arbeit und Technik,
Gelsenkirchen, 1999.
Anxo, D. and Tanemar, L. (1996), ‘Subventioner till hemtjänstesektorn, erfarenheter från
Danmark och Frankrike’, Invandrapolitiska kommittén, SOU 1996:55, bilaga 1. Stockholm,
1996.
Appelbaum, E. (1990), ‘Unbalanced growth and the US employment expansion’, Structural
Change and Economic Dynamics, Vol. 1, No 1, pp. 91–101.
Appelbaum, E. and Schettkat, R. (1995), ‘Employment and productivity in industrialised
economies’, International Labour Review, Vol. 134, No 4–5, pp. 605–623.
Ashton, D. and Green, F. (1996), Education, training and the global economy, Edward Elgar.
Atkinson, A. (1995), ‘The welfare state and economic performance’, National Tax Journal,
48, pp. 171–198.
Aukrust, O. (1977), ‘Inflation in the open economy: a Norwegian model’, in Krause, L. B. and
Salant, W. S. (eds), Worldwide inflation, The Brookings Institution, Washington DC,
pp. 107–153.
Bäcker, G. and Hanesch, W. (1997), ‘Kombi-Lohn: Kein Schlüssel zum Abbau der
Arbeitslosigkeit!’, in WSI Mitteilungen, 50, October, pp. 701–712.
Baethge, M. (1999), ‘Beschäftigung und Tertiarisierung’ (‘Employment and tertiarisation’), in
Bullinger, H.-J. (ed.), Dienstleistungen — Innovation für Wachstum und Beschäftigung.
Herausforderungen des internationalen Wettbewerbs, Gabler, Wiesbaden, pp. 359–363.
Balassa, B. (1964), ‘The purchasing power parity doctrine: a reappraisal’, Journal of Political
Economy, Vol. 72(2), pp. 584–596.
Barcet, A. and Bonamy, J. (1994), ‘Qualité et qualification des services’, in De Bandt, J. and
Gadrey, J. (eds), Relations de service, marchés de services, CNRS Editions, pp. 153–174.
Baudry, B. (1992), ‘Contrat, autorité, confiance: la relation de sous-traitance est-elle
assimilable à la relation d’emploi?’, Revue Économique, Vol. 43, No 5, September,
pp. 871–894.
Baudry, B. (1995), L’économie des relations inter-entreprises, La Découverte, coll. ‘Repères’.
Baumol, W. J. (1967), ‘Macroeconomics of unbalanced growth: the anatomy of urban crisis’,
American Economic Review, Vol. 57, pp. 415–426.
304
Bibliography
Baumol, W. J. (1986), ‘Productivity growth, convergence and welfare: what the long-run data
show’, American Economic Review, Vol. 76, pp. 1072–1085.
Baumol, W. J., Blackman, S. A. B. and Wolff, E. N. (1985), ‘Unbalanced growth revisited:
asymptotic stagnancy and new evidence’, American Economic Review, Vol. 75, pp. 806–817.
Baumol, W. J. (1989), Productivity and American leadership: the long view, Cambridge, MIT
Press.
Baumol, W. J., Blackman, S. A. B. and Wolff, E. N. (1992), Productivity and American
leadership: the long view, Cambridge, MIT Press.
Baumol, W. J. and Bowen, W. G. (1966), Performing arts — The economic dilemma,
Twentieth Century Found, Cambridge, Mass.
Becker, G. (1965), ‘A theory of the allocation of time’, Economic Journal, 75, pp. 493–517.
Becker, I. (1998), Vergleich und Bewertung alternativer Grundsicherungskonzepte (working
papers for the EVS project ‘Personelle Einkommensverteilung in der Bundesrepublik
Deutschland’, No 18), Frankfurt am Main University, Institut für Konjunktur, Wachstum und
Verteilung.
Begg, I. (1993), ‘The service sector in regional development’, in Regional Studies, No 8,
pp. 817–825.
Beise, M., Gehrke, B. and Legler, H. (1999), ‘Attraktivität Deutschlands und seiner Regionen
für Forschungs- und Entwicklungsaktivitäten’ (‘The attraction of Germany and its regions for
research and development’), in Informationen zur Raumentwicklung, No 1, pp. 31–44.
Beise, M. and Gehrke, B. (1998), Zur regionalen Konzentration von Innovationspotentialen
in Deutschland (On the regional concentration of innovation potential in Germany),
Documentation No 98-09, ZEW, Mannheim.
Bell, D. (1999), The coming of post-industrial society. A venture in social forecasting, Basic
Books, New York (with a new foreword).
Bell, D. (1975) (1973), Die nachindustrielle Gesellschaft, Campus Verlag, Frankfurt am
Main.
Bellmann, L. (1998), ‘Struktur der Arbeitslosigkeit und Sozialhilfebedürftigkeit’, in Sozialer
Fortschritt, 47, August, pp. 207–209.
Bellmann, L. and Jackman, R. (1996), ‘The impact of labour market policy on wages,
employment and labour market mismatch’, in O’Reilly, G. and Klaus, J. S. (eds),
International handbook of labour market policy and evaluation, Elgar, pp. 725–746.
305
The job creation potential of the service sector in Europe
Bender, S., Kaltenborn, B., Rudolph, H. and Walwei, U. (1999), ‘Förderung eines
Niedriglohnsektors: Die Diagnose stimmt, die Therapie noch nicht’, in IAB Kurzbericht, No 6.
Benhamou, F. (1997), L’économie de la culture, La Découverte, coll. ‘Repères’, Paris.
Benhamou, F. (forthcoming), ‘The opposition between two models of labour market
adjustment — The case of the audio-visual and performing arts activities in France and Great
Britain over a ten-year period’, Journal of Cultural Economics.
Benito, L. (1995), Tourisme et festivals, mémoire de DEA, mimeo, Aix Marseille University.
Benus, J., Wood, M. and Grover, N. (1994), A comparative analysis of the Washington and
Massachussets UI self-employment demonstrations, Abt, Bethesda.
Bernard, A. and Jones, C. (1996), ‘Productivity across industries and countries: time series
theory and evidence’, The Review of Economics and Statistics, Vol. 78(1), pp. 135–146.
Bernstein, P. L. (1996), Against the gods. The remarkable story of risk, John Wiley & Sons,
New York et al.
Beyers, W. B. (1991), ‘Trends in the producer services in the USA: the last decade’, in
Daniels, P. W. (ed.), Services and metropolitan development: international perspectives,
Routledge, London and New York, pp. 146–172.
Bittner, S. and Weinkopf, C. (1999), ‘Geringfügige Beschäftigung in Deutschland — Daten,
Probleme und Kommentierung der aktuellen Debatte’, in Zukunft im Zentrum (ed.),
Innovative Dienstleistungen für private Haushalte, Berlin, pp. 70–77.
Blanchflower, D. and Burgess, S. (1996), ‘New technology and jobs: comparative evidence
from a two country study’, LSE Centre for Economic Performance Discussion Paper No 285.
Blanchflower, D. G. and Oswald, A. J. (1994), The wage curve, Cambridge, Mass., MIT Press,
London.
Blien, U. (1996), ‘Die ‘‘Lohnkurve’’: Eine Einführung zu den Workshop-Beiträgen’, in
Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, 29(3), pp. 455–459.
Bofinger, P. and Fasshauer, S. (1998), ‘Reduzierung der Sozialabgaben statt Kombilohn’, in
Wirtschaftsdienst, 78, No IX, pp. 519–528.
Boitani, A. and Pellegrini, G. (1997), ‘Lo Sviluppo di Nuove Attività Produttive: i Servizi’, in
Ciocca, P. (ed.), Disoccupazione di Fine Secolo — Studi e Proposte per l’Europa, Bollati
Boringhieri, Torino, pp. 203–267.
306
Bibliography
Boje, T. (1995), Women and the European employment rate: the causes and consequences of
variations in female activity and employment patterns, report to the European Commission,
EC network on the situation of women in the labour market, Manchester School of
Management, UMIST.
Bopp, C. (1996), ‘Zukunftsmärkte für Dienstleistungen — das Feld Handel, Distribution,
Konsum und Freizeit’, in IFO Schnelldienst, 49, No 25/26, pp. 30–32.
Boulding, K. E. (1978), Ecodynamics — A new theory of societal evolution, Sage
Publications, Beverly Hills and London.
Bouvet, P. (1996), ‘Les salaires des ‘‘vedettes’’ du sport professionnel par équipe: une
application privilégiée de la théorie du salaire d’efficience’, Revue française d’économie,
XI(4), pp. 119–144.
Bowen, D. E. and Jones, G. R. (1986), ‘Transaction cost analysis of service organisationcustomer exchange’, Academy of Management Review, Vol. 11, No 2, pp. 428–441.
Brake, K. (1993), ‘Die räumliche Struktur der Dienstleistungsökonomie oder Warum gibt es
keine Dezentralisierung?’ (‘The geographical structure of the service economy, or why is there
no decentralisation?’), in Häußermann, H. and Siebel, W. (eds), New York — Strukturen einer
Monopole, Suhrkamp, Frankfurt am Main, pp. 91–107.
Brake, K. (1994), ‘Dezentrale Konzentration — Zum Verhältnis von Leitbild und
Standorttendenzen’ (‘Decentral concentration — On the relationship between models and
locational trends’), in Informationen zur Raumentwicklung, No 7/8, pp. 481–530.
Branson, W. H. and Myhrman, J. (1976), ‘Inflation in open economies: supply-determined
versus demand-determined models’, European Economic Review, Vol. 7(1), pp. 15–34.
Bronstein, A. S. ‘Temporary work in Western Europe: threat of complement to permanent
employment’, International Labour Review, Vol. 130, pp. 291–310.
Brusco, S. (1992), ‘Small firms and the provision of real services’, in Pyke, F. and
Sengenberger, W. (eds), Industrial districts..., loc. cit., pp. 177–196.
Bullinger, H.-J. (1998), Dienstleistung 2000plus. Zukunftsreport Dienstleistungen in
Deutschland, Fraunhofer IRB-Verlag, Stuttgart.
Bundesministerium für Bildung und Forschung (BMBF) (1999), Zur technologischen
Leistungsfähigkeit Deutschlands (On Germany’s technical efficiency), Zusammenfassender
Endbericht, 1998, BMBF, Bonn.
Burchell, B. (1996), ‘Gender segregation, size of workplace and the public sector’, Gender,
Work and Organisation, 3(4), pp. 227–240.
307
The job creation potential of the service sector in Europe
Burton, J. (1999), Inflexible friend: the new flexible economy and professional services, Adam
Smith Institute.
Cabinet Ithaque, Cabinet Themis (1997), Le spectacle vivant, La documentation française,
Paris.
Calmfors, L. (1977), ‘Inflation in Sweden’, in Krause, L.B. and Salant, W.S. (eds), Worldwide
Inflation, Washington DC, The Brookings Institution, pp. 107–153.
Camagni, R. (1991), Innovation networks — Spatial perspectives, Belhaven Press,
London/New York.
Campbell, M. and Daly, M. (1992), ‘Self-employment: into the 1990s’, Employment Gazette,
June, pp. 269–292.
Capelli, P., Bassi, L., Katz, H., Knoke, D., Osterman, P. and Useem, M. (1997), Change at
work, Oxford University Press, Oxford.
Casey, B. et al. (1995), ‘Employment in the cultural sector’, Cultural Trends, 25, pp. 88–97.
Castells, M. (1996), The rise of the network society, Blackwell, Oxford.
CCE (1991), Panorama de l’industrie communautaire 1991.
CCE (1997), Panorama de l’industrie communautaire 1997.
CEC (1997), Employment in Europe 1997, Office for Official Publications of the European
Communities, Luxembourg.
CERC (1994), Social welfare and economic activity of women in Europe, V/2184/94-EN,
European Commission, Brussels.
Cette, D. et al. (1998), ‘Emplois de proximités’, Conseil d’analyse économique, Report No 12,
La documentation française, Paris.
Christopherson, S. and Storper, M. (1989), ‘The effects of flexible specialisation on industrial
politics and the labour market’, Industrial and Labor Relations Review, 42(3), pp. 331–347.
Clark, C. (1940), The conditions of economic progress, Macmillan, London.
Cohen, W. M. and Levinthal, D. A. (1990), ‘Absorptive capacity: a new perspective on
learning and innovation’, Administrative Science Quarterly, Vol. 35, March, pp. 128–152.
Cohendet, P. (1994), ‘Relations de service et transfert de technologie’, in De Bandt, J. and
Gadrey, J. (eds), Relations de service, marchés de services, CNRS Editions, pp. 201–213.
Commission Européenne (1997), L’emploi en Europe, 1997.
308
Bibliography
Conseil de l’Europe (1996), Annuaire statistique, Observatoire européen de l’audiovisuel,
Brussels.
Cornetz, W. (1991), ‘Ökonomische Aspekte des dienstleistungsorientierten Strukturwandels’
(‘Economic aspects of service-related structural change’), in Littek, W., Heisig, U. and
Gondak, H.-D. (eds), Dienstleistungen. Strukturveränderungen, Beschäftigungsbedingungen
und Interessenlagen, Edition Sigma, Berlin, pp. 35–52.
Cornetz, W. and Schäfer, H. (1998), ‘Hat Deutschland den Dienstleistungsrückstand
gegenüber den USA aufgeholt?’, Wirtschaftsdienst, 78, No VII, pp. 418–425.
Coyle, D. (1997), The weightless world, Capstone Books.
Cully, M. et al. (1999), Britain at work, Routledge.
Daniels, P. W. (1985), Service industries — A geographical appraisal, Methuen, London and
New York.
Daniels, P. W. (1991), ‘Service sector restructuring and metropolitan development: processes
and prospect’, in Daniels, P. W. (ed.), Services and metropolitan development: international
perspectives, Routledge, London and New York.
Daniels, P. W. (1999), ‘From global to local — Export of services or servicing exports?’, in
Bullinger, H.-J. (ed.), Dienstleistungen — Innovation für Wachstum und Beschäftigung.
Herausforderungen des internationalen Wettbewerbs, Gabler, Wiesbaden, pp. 88–105.
DARES (1998), ‘Structure des emplois en 1995’, Les dossiers de la DARES, No 3, December.
De Bandt, J. (1995), Services aux entreprises: informations, produits, richesses, Economica.
De Bandt, J. (1996), ‘Coopération, accords inteentreprises, concurrence’, in Ravix, J.-L. (ed.),
Coopérations entre les entreprises et organisation industrielle, CNRS Editions, pp. 195–229.
De Bandt, J. and Gadrey, J. (1994), Relations de service, marchés de services, CNRS Editions.
Deitelhoff, N. (1999), ‘Anmerkungen zu Rahmenbedingungen des öffentlich geförderten
haushaltsnahen Dienstleistungsbereichs’, in Zukunft im Zentrum (ed.), Innovative
Dienstleistungen für private Haushalte, Berlin, pp. 57–59.
Delaunay, J.-C. and Gadrey, J. (1987), Les enjeux de la société de service, Presses de la
Fondation nationale des sciences politiques.
Delors, J. (1996), ‘Development — A new model’, European Labour Forum, winter.
309
The job creation potential of the service sector in Europe
Deven, F., Inglis, S., Moss, P. and Petrie, P. (1997), State of the art review on the reconciliation
of work and family life for men and women and the quality of care services, final report for the
European Commission, Equal Opportunities Unit.
Dex, S. et al. (1998), ‘Workers’ strategies in uncertain labour markets: analysis of the effects
of casualisation in the television industry’, paper given at the Work Employment and Society
Conference, University of Cambridge, September.
Doeringer, P. and Piore, M. (1971), Internal labor markets and manpower analysis, Heath,
Washington DC.
Döhl, V. and Sauer, D. (1995), ‘Neue Unternehmensstrategien und regionale Entwicklung’
(‘New enterprise strategies and regional development’), Jahrbuch sozialwissenschaftliche
Technikberichterstattung 1995, Edition Sigma, Berlin, pp. 103–157.
Dosi, G. and Marengo, L. (1994), ‘Some elements of an evolutionary theory of organisational
competences’, in England R. W. (ed.), Evolutionary concepts in contemporary economics,
University of Michigan Press, Ann Arbor, pp. 157–178.
Dreger, C., Kempe, W., Kolb, J. and Trabert, L. (1998), ‘Was bringt ein Kombilohn?’,
Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, 31, No 4, pp. 705–714.
Drennan, M. P. (1992), ‘Gateway cities: the metropolitan sources of US producer service
exports’, Urban Studies, Vol. 29, No 2, pp. 217–235.
DTI (1999), Wired world — Work in the knowledge-driven economy, DTI Futures Unit,
August.
Dupuy, M. and Schweitzer, M. E. (1994), ‘Are service sector jobs inferior?’, Economic
Commentary, Federal Reserve Bank of Cleveland, 1 February.
Eberts, R. W. and Groshen E. L., (1990), ‘Service jobs versus manufacturing jobs: How do
they compare?’, Economic Development Commentary, Vol. 14, No 2, summer.
Edeltruat, H. and Walwei, U. (1999), The change in employment forms — Empirical results
and first explanatory approaches, Institute for Employment Research, No 34.
Edgren, G., Faxén, K.-O. and Odhner, C. E. (1973), Wage formation and the economy, Allen
& Unwin, London.
Elfring, T. (1988), Service sector employment in advanced economies, Avebury, Aldershot.
Erickcek, G. and Houseman, S. (1997), Summary findings of a report on temporary, part-time,
and contract employment in the United States, W. E. Upjohn Institute for Employment
Research, Kalamazoo, MI.
310
Bibliography
Escott, K. and Whitfield, D. (1995), ‘The gender impact of CCT in local government’, Equal
Opportunities Commission Research Series No 12, Equal Opportunities Commission,
Manchester, UK.
Esping-Anderson, G. (1990), The three worlds of welfare capitalism, Polity Press, London.
Esping-Andersen, G. (1996), Welfare states in transition, Sage Publications, London.
Esping-Andersen, G. (1999), Social foundations of post-industrial economies, Oxford
University Press, Oxford.
European Commission (1997), The Social Economy Unit of the European Commission.
Europäische Kommission (1995), Lokale Initiativen zur wirtschaftlichen Entwicklung und
Beschäftigung, Luxembourg.
Europäische Kommission (1999), ‘Entschließung des Rates zu den Leitlinien für
beschäftigungspolitische Maßnahmen der Mitgliedsstaaten 1999’, Bulletin der Europäischen
Kommission, No 1/2, pp. 2–12.
European Commission (1998), Employment rates report.
European Commission (1998), ‘Modernising public employment services to support the
European employment strategy’, Commission communication, draft, Brussels, 9 November
1998, 19 pp.
Evers, A. (1995), ‘Von Dienstleistungen zu Geldleistungen?’, in Gräbe, S. (ed.), Private
Haushalte und neue Arbeitsmodelle, Campus, Frankfurt am Main/New York, pp. 119–140.
Evers, A. (1999), ‘Haushaltsnahe Dienstleistungen als innovativer Angebots- und
Beschäftigungsfaktor’, Sozialer Fortschritt, 48, May, pp. 105–110.
Fagan, C. (1997), ‘Absent men and juggling women: gender, households and working-time
schedules in Britain’, Doctoral thesis, University of Manchester, UK.
Fagan, C. and Rubery, J. (1996), ‘Transitions between family formation and paid
employment’, in Schmid, G., O’Reilly, J. and Schömann, K. (eds) International handbook of
labour market policy and evaluation, Edward Elgar, Cheltenham.
Fagan, C., Plantenga, J. and Rubery, J. (1995), ‘Part-time work and inequality? Lessons from
the Netherlands and the UK’, in Lapeyre, J. and Hoffman, R. op. cit.
Faust, K. et al. (1999), ‘Tertiarisierung und neue Informations- und
Kommunikationstechnologien’, IFO-Schnelldienst, No 29, pp. 23–34.
311
The job creation potential of the service sector in Europe
Felli, E., Rosati, F. C. and Tria, G. (1995), The service sector: productivity and growth,
proceedings of the International Conference held in Rome, Italy, May 1993, Physica,
Heidelberg.
Felstead, A. et al. (1998), ‘Insecurity at work’, New Economy, Vol. 5, No 3, Institute of Public
Policy Research, September.
Ferber, M. and Waldfogel, J. (1996), ‘Contingent work: blessing and/or curse’, Radcliffe,
pp. 1–53.
Filer, R. K. (1986), ‘The ‘‘starving’’ artist — Myth or reality? Earnings of artists in the United
States’, Journal of Political Economy, 96, pp. 56–75.
Finger, D. (1997), ‘Dienstleistungsschecks in Europa — Ein Modell für Deutschland?’,
Discussion Paper FS I 97-201 — English version available), Wissenschaftszentrum Berlin für
Sozialforschung, Berlin.
Fölster, K. (1999), ‘Paid domestic work in Germany’, mimeo.
Fontaine, C. (1987), L’expansion des services: un quart de siècle en France et dans le monde
développé, Rexervices.
Foss, N. J. (1993), ‘Theories of the firm: contractual and competence perspectives’, Journal
of Evolutionary Economics, Vol. 3, pp. 127–144.
Fourastié, J. (1969) (1963), Die große Hoffnung des zwanzigsten Jahrhunderts (The great
hope of the twentieth century), second edition (first edition, 1954), Bund, Cologne.
Fourastié, J. and Schneider, J. (1989), Warum die Preise sinke — Produktivität und Kaufkraft
seit dem Mittelalter, Campus, Frankfurt am Main.
Frank, R. H. and Cook, P. J. (1995), The winner-take-all society, The Free Press, New York
et al.
Freeman, C. (1996), ‘Technical change and unemployment’, in Philpott, J. (ed.), Working for
full employment, Routledge.
Freeman, R. (1995), ‘The limits of wage flexibility to curing unemployment’, Oxford Review
of Economic Policy, Vol. 11(1), pp. 63–72.
Freeman, R. (1996), ‘Does globalisation threaten low-skilled western workers’, in Philpott, J.
(ed.), Working for full employment, Routledge.
312
Bibliography
Freeman, R. and Schettkat, R. (1999), Differentials in service industry employment growth:
Germany and the US in the comparable German-American structural database, report
prepared for the European Commission, Directorate-General for Employment and Social
Affairs, mimeo.
Frisch, H. (1983), Theories of inflation, CUP, Cambridge.
Fuchs, V. (1968), The service economy, NBER, Columbia University Press, New York.
Fujita, M., Krugman, P. and Venables, A. J. (1999), The spatial economy — Cities, regions,
and international trade, MIT Press, Cambridge, Mass., and London.
Gadrey, J. (1990), ‘Rapports sociaux de service: une autre régulation’, Revue Economique,
Vol. 41, No 1, January, pp. 49–69.
Gadrey, J. (1992), L’économie des services, La Découverte, coll. ‘Repères’.
Gadrey, J. (1996), Services: la productivité en question, Desclée de Brouwer, Paris, coll.
‘Sociologie économique’.
Gadrey, J. and Jany-Catrice, F. (1997), ‘L’emploi commercial aux Etats-Unis et en France et
les difficultés des comparaisons internationales’, La Revue de l’IRES, No 25, autumn.
Gadrey, J. and Jany-Catrice, F. (1998), ‘Emploi et marchés du travail d’un secteur: quelles
comparaisons internationales?’, Economie et Statistique, No 318, p. 8.
Gadrey, J., Jany-Catrice, F. and Ribault, T. (1998), ‘L’emploi dans le commerce de détail
japonais, une comparaison avec la France’, Formation Emploi, No 61, January–March.
Ganz, W. (1999), ‘Wachstumsfaktor Dienstleistungen’, in Koller, M. (ed.), Innovations- und
Wachstumspotentiale neuer Dienstleistungen, Beiträge zur Arbeitsmarkt- und
Berufsforschung (BeitrAB 222), Nürnberg, pp. 41–66.
Genosko, J. (1999), Netzwerke in der Regionalpolitik (Networks in regional policy), Schüren
Verlag, Marburg.
Gershuny, J. (1983), Social innovation and the division of labour, Oxford University Press,
Oxford.
Gershuny, J. and Miles, I. (1983), The new service economy, Frances Pinter, London.
Gershuny, J., Godwin, M. and Jones, S. (1994), ‘The domestic labour revolution: a process of
lagged adaptation’, in Anderson, M., Bechhofer, F. and Gershuny, J. (eds) (1994a), The social
and political economy of the household, Oxford University Press, Oxford.
313
The job creation potential of the service sector in Europe
Gerster, F. and Deubel, I. (1999), ‘Arbeit muß sich lohnen!’, Wirtschaftsdienst, 79, No I,
pp. 39–43.
Giannelli, G. C. and Esping-Andersen, G. (1990), ‘Labour costs and employment in the
service economy’, Labour, Vol. 4(3), pp.107–132.
Giarini, O. and Stahel, W. R. (1989), The limits to certainty — Facing risks in the new service
economy, Kluwer Academic Publishers, Dordrecht, Boston, London.
Goldberg, M. A., Helsley, R. W. and Levi, M. D. (1991), ‘The growth of international financial
services and the evolution of international financial centres: a regional and urban economic
approach’, in Daniels, P. W. (ed.), Services and metropolitan development: international
perspectives, Routledge, London and New York.
Gornig, M., Ring, P. and Reiner, S. (1999), ‘Strategische Dienstleistungen in Hamburg: Im
Städtevergleich gut positioniert’ (‘Strategic services in Hamburg: a good comparative
position’), DIW Wochenbericht, No 4.
Grabher, G. (1994), Lob der Verschwendung (In praise of waste), Edition Sigma, Berlin.
Greffe, X. (1996), ‘La contribution du secteur culturel au développement de l’emploi dans
l’Union Européenne’, report for DG V, European Commission.
Gregg, P. and Wadsworth, J. (1996), The polarisation of work, Employment Policy Institute
Employment audit, November.
Gregson, N. and Lowe, M. (1994), Servicing the middle class: class, gender and waged
domestic labour in contemporary Britain, Routledge, London.
Gries, T. and Birk, A. (1999), ‘Die amerikanische Dienstleistungsgesellschaft — ein Modell
für Deutschland?’, Wirtschaftsdienst, 79, No V, pp. 300–306.
Grillches, Z. (1987) Productivity in the New Palgrave: A dictionary of economics, Macmillan
Press, London.
Grillches, Z. (1991) Output measurement in the service sectors, National Bureau of Economic
Research, Conference on Research in Income and Wealth, University of Chicago Press,
Chicago and London.
Groshen, E. L. (1987), ‘Can services be a source of export-led growth? Evidence from the
Fourth District’, Economic Review, Federal Reserve Bank of Cleveland, Quarter 3.
Groshen, E. L. and Robertson, L. (1993), ‘Are the Great Lake cities becoming service
centers’, Economic Commentary, Federal Reserve Bank of Cleveland, 1 June.
314
Bibliography
Gurgand, M. and Menger, P. M. (1996), ‘Work and compensated unemployment in the
performing arts — Exogenous and endogenous uncertainty in artistic labour markets’, in
Ginsburgh, V. and Menger, P. M. (eds) (1996), Economics of the arts, Elsevier, Amsterdam,
pp. 347–381.
Haak, C. and Schmid, G. (1999), ‘Arbeitsmärkte für Künstler und Publizisten — Modelle
einer zukünftigen Arbeitswelt?’, Discussion Paper P 99-506, Wissenschaftszentrum Berlin für
Sozialforschung.
Haisken-DeNew, J., Horn, G. A., Schupp, J. and Wagner, G. (1996), ‘Keine
Dienstleistungslücke in Deutschland’, in DIW Wochenbericht, 63, No 14, pp. 221–226.
Haller, S. (1999), ‘Die Beurteilung von Dienstleistungsqualität im Wertschöpfungsverbund’
(‘Assessment of service quality in value added alliances’), in Bullinger, H.-J. (ed.),
Dienstleistungen..., loc. cit., pp. 545–561.
Hansen, N. (1990), ‘Does producer service induce regional economic development’, Journal
of Regional Science, Vol. 30, No 4, p. 467.
Häußermann, H. (ed.) (1992), Ökonomie und Politik in alten Industrieregionen Europas —
Probleme der Stadt- und Regionalentwicklung in Deutschland, Frankreich, Großbritannien
und Italien (Economics and politics in Europe’s old industrial regions — Problems of urban
and regional development in Germany, France, Great Britain and Italy), Birkhäuser,
Basel/Boston/Berlin.
Häußermann, H. and Siebel, W. (1995), Dienstleistungsgesellschaften (Service societies),
Suhrkamp, Frankfurt am Main.
Heinze, R. and Schmid, J. (1994), Industrieller Strukturwandel und die Kontingenz politischer
Steuerung: Mesokorporatistische Strategien im Vergleich (Structural change in industry and
the contingency of political control: meso-corporatist strategies in comparison) (SIT-wp-294), Forschungsstelle für Sozialwissenschaftliche Innovations- und Technologieforschung an
der Ruhr-Universität Bochum.
Henrekson (1998), ‘En ond cirkel för tjänstesektorn, Arbetsmarknad & Arbetsliv, årg 4, nr 2,
sommaren 1998.
Heskett, J. L., Sasser, W. Earl jr and Schlesinger, L. A. (1997), The service profit chain — How
leading companies link profit and growth to loyalty, satisfaction and value, The Free Press,
301 pp.
Hicks, J. R. (1932), The theory of wages, Macmillan, London.
Hild, R., Hofmann, H., Ochel, W. and Wilhelm, M. (1999), ‘Unternehmensbezogene
Dienstleistungen: Deutschlands Position im internationalen Wettbewerb’ (‘Business services:
Germany’s competitive position’), IFO Schnelldienst, No 10/11, pp. 3–19.
315
The job creation potential of the service sector in Europe
Hill, T. P. (1977), ‘On goods and services’, Review of Income and Wealth, No 4, pp. 315–338.
Hochschild, A. (1990), The second shift, Piatkus, London.
Hodgson, G. M. (1998), ‘Competence and contract in the theory of the firm’, Journal of
Economic Behavior and Organization, Vol. 35, No 2, pp. 179–201.
Hogan, H. (1999), ‘Strategie für neue Beschäftigung – aus der Arbeitslosigkeit in den ersten
Arbeitsmarkt’, in Koller, M. (ed.), Innovations- und Wachstumspotentiale neuer
Dienstleistungen, Beiträge zur Arbeitsmarkt- und Berufsforschung (BeitrAB 222), Nürnberg,
pp. 37–40.
Hollard, M. and Ruffieux, B. (1996), ‘Cohérence et pertinence des frontières d’entreprises’,
in Ecosip, Cohérence, pertinence et évaluation, Economica, pp. 159–176.
Horrell, S. (1991). ‘Working-wife households: inside and outside the home’, unpublished PhD
thesis, University of Cambridge, UK.
Horrell, S. (1994), ‘Household time allocation and women’s labour force participation’, in
Anderson, M., Bechhofer, F. and Gershuny, J. (eds) (1994a), The social and political economy
of the household, Oxford University Press, Oxford.
Houseman, S. (1996), Temporary, part-time, and contract employment in the United States: a
report on the W.E. Upjohn Institute’s Employer Survey on Flexible Staffing Policies, W.E.
Upjohn Institute for Employment Research, Kalamazoo, MI.
Hummel, M. (1999), ‘Nationale und internationale Beschäftigungspotentiale im Bereich
unternehmensbezogener Dienstleistungen’ (‘National and international employment potential
in business services’), in Bullinger, H.-J. (ed.), Dienstleistungen..., loc. cit., pp. 381–387.
Hutton, W. (1999), ‘Partnership, empowerment and the new deal for communities’, EPI
Policy Monitor, February.
Huws, U. (1995), ‘Teleworking: follow-up to the White Paper’, Social Europe, Supplement
3/95, Office for Official Publications of the European Communities, Luxembourg, pp. 1–96.
Incomes Data Services (1999), Pay and conditions in call centres, September. See also Mitial
(1999), UK and Eire call centre study.
Inman, R. P. (ed.) (1985), Managing the service economy — Prospects and problems,
Cambridge University Press, Cambridge.
INSEE (1999), ‘L’emploi dans les services marchands’, INSEE Synthèses, No 24, April.
INSEE, ‘Structure des emplois au 31.12, INSEE Résultats’, Emploi-Revenus, different years.
316
Bibliography
Jaeger, C. and Durrenberger, G. (1991), ‘Services and counterurbanization: the case of Central
Europe’, in Daniels, P. W.(ed.), Services and metropolitan development: international
perspectives, Routledge, London and New York.
Juhn, C., Murphy, K. M. and Pierce, B. (1993), ‘Wage inequality and the rise in returns to
skills’, Journal of Political Economy, Vol. 101(3), pp. 410–442.
Kaiser, U. and Voß, K. (1999), ‘Do business-related services really lag behind manufacturing
industries in the business cycle? Empirical evidence on the lead/leg relationship between
business-related services and manufacturing industries for Germany, ZWE Discussion Paper
No 3, ZEW, Mannheim.
Kaltenborn, B. (1998), ‘Grundsicherungskonzepte von Parteien und Verbänden’,
Gewerkschaftliche Monatshefte, 49, September, pp. 587–594.
Katouzian, M. (!970), ‘The development of the service sector: a new approach’, Oxford
Economic Papers, pp. 362–382.
Keep, E. and Mayhew, K. (1998), ‘Was Ratner right? Product market and competitive
strategies and their links with skills and knowledge’, Employment Policy Institute Economic
Report, Vol. 12, No 3, April.
Kendrick, J. (1973), Long-term economic growth 1860–1970, US Bureau of Census,
Washington DC, June 1973.
Kilper, H., Latniak, E., Rehfeld, D.and Simonis, G. (1994), Das Ruhrgebiet im Umbruch —
Strategien regionaler Verflechtung (Upheaval in the Ruhr Valley —Strategies of regional
interpenetration), Leske + Budrich, Opladen.
Klös, H.-P. (1997), ‘Dienstleistungslücke und Niedriglohnsektor in Deutschland’, IW-Trends,
77, No 3, pp. 33–59.
Kommission für Zukunftsfragen der Freistaaten Bayern und Sachsen (1997), Erwerbstätigkeit
und Arbeitslosigkeit in Deutschland, Bonn.
Kritikos, A. and Wießner, F. (1999), ‘Hilfen für Existenzgründer — Bessere Fördertechnik in
den Vereinigten Staaten’, IAB-Kurzbericht, Ausgabe No 15, 1 October 1999.
Krueger, A. B. and Pischke, J. S. (1997), ‘Observations and conjectures on the US
employment miracle’, NBER Working Paper No 6146.
Krugman, P. (1995), ‘Technology, trade and factor prices’, NBER Working Paper No 5355.
Krugman, P. (1996), ‘Domestic distortions and the deindustrialization hypothesis’, NBER
Working Paper No 5473.
317
The job creation potential of the service sector in Europe
Kuznets, S. (1996), Modern economic growth: rate structure and spread, Yale University
Press, New Haven.
Lammers, K. (1999), ‘Räumliche Wirkungen der Globalisierung in Deutschland’ (‘Spatial
effects of globalisation in Germany’), Informationen zur Raumentwicklung, No 1, pp. 9–18.
Läpple, D. (1989), ‘Neue Technologien in räumlicher Perspektive’ (‘New technologies from
a spatial perspective’), Informationen zur Raumentwicklung, No 4, pp. 213–226.
Läpple, D. (1994), ‘Die Teilökonomie einer Großstadt in einer Phase strukturellen Umbruchs
— Das Beispiel Hamburgs’ (‘The partial economy of a city undergoing structural change —
The example of Hamburg’), Beiträge zur Arbeitsmarkt- und Berufsforschung, Vol. 184,
pp. 106–139.
Leadbeater, C. (1999), Living on thin air — The new economy, Viking Books.
Lennartz, D. (1999), ‘Zeitgemäße Strategien der beruflichen Bildung für Dienstleistungen’
(‘Up-to-date strategies of vocational training for services’), Bullinger, H.-J. (ed.),
Dienstleistungen..., loc. cit., pp. 460–473.
Levine, R. ‘Financial development and economic growth’, Journal of Economic Literature,
35, pp. 688–726.
Levitt, B. and March, J. G. (1988), ‘Organisational learning’, Annual Review of Sociology,
Vol. 14, pp. 319–340.
Licht, G., Hipp, C., Kukuk, M. and Münt, G. (1997), ‘Innovationen im Dienstleistungssektor
— Empirischer Befund und wirtschaftspolitische Konsequenzen’ (‘Innovation in the service
sector — Empirical diagnosis and consequences for economic policy’), Schriftenreihe des
ZEW, Vol. 24, Nomos, Baden-Baden.
Lichtblau, K. and Rhein, A. (1993), ‘Regionaler Strukturwandel in Westdeutschland 1980/92’
(‘Structural change in the regions of West Germany, 1980/92’), IW-Trends, No 4, pp. 57–74.
Lindbeck, A. (1979), ‘Imported and structural inflation and aggregate demand: the
Scandinavian model reconstructed’, in Lindbeck, A. (ed.) Inflation and employment in open
economies, Amsterdam, North Holland, pp. 13–40.
Lucas, R. (1988), ‘On the mechanics of economic development’, Journal of Monetary
Economics, 22, pp. 3–42.
Lundvall, B. A. (1992), ‘Relations entre utilisateurs et producteurs, systèmes nationaux
d’innovation et internationalisation’, in Foray, D. and Freeman, C. (eds), Technologie et
richesse des nations, Economica, pp. 355–388.
318
Bibliography
Lundvall, B. A. (1988), ‘Innovation as an interactive process: from user-producer interaction
to the national system of innovation’, in Dosi, G., Freeman, C., Nelson, R., Silverberg, G. and
Soete, L. (eds), Technical change and economic theory, Pinter Publishers, pp. 349–369.
Maddison, A. (1980), ‘Economic growth and structural change in advanced countries’, in
Leveson and Wheeler (eds), Western economies in transition: structural change and
adjustment policies in industrial countries, Westview Press, Boulder.
Maddison, A. (1986), ‘Growth and slowdown in advanced capitalist economies: techniques of
quantitative assessment’, Journal of Economic Literature, 25, pp. 649–706.
Madrick, J. (1999), ‘How new is the new economy?’, The New York Review of Books, Vol.
XLVI, No 14, September 1999.
Maier, F. and Rapp, Z. (1995), Women and the European employment rate: the causes and
consequences of variations in female activity and employment patterns in Germany, report to
the European Commission, EC network on the situation of women in the labour market,
Manchester School of Management, UMIST.
Maier, F., Quack, S., Martschink, A. and Rapp, Z. (1996), Trends and prospects for women’s
employment in Germany in the 1990s, report to the European Commission, EC network on the
situation of women in the labour market, Manchester School of Management, UMIST.
Manacorda, M. and Petrongolo, B. (1999), ‘Skill mismatch and unemployment in OECD
countries’, Economica, pp. 181–207.
Marimon, R. and Zilibotti, F. (1996), ‘Actual versus virtual employment: is Spain different?’,
CEPR Discussion Paper No 1427.
Martinelli, F. (1991), ‘Producer services’ location and regional development’, in Daniels, P.
W. and Moulaert, F. (eds), The changing geography of advanced producer services, Belhaven
Press, London and New York.
McGregor, A. (1997), The contribution of the Third Sector to the economies of disadvantaged
communities, Community Enterprise in Scotland.
Meadows, P. (1999), The flexible labour market — Implications for pension provision,
National Association of Pension Funds, May 1999.
Meager, N. (1994), ‘Self-employment schemes for the unemployed in the European
Community — The emergence of a new institution, and its evaluation’, in Schmid, G. (ed.),
Labour market institutions in Europe, Armonk, New York and London, pp. 183–242.
Meager, N. (1996), ‘From unemployment to self-employment’, in Schmid, G., O’Reilly, J. and
Schöman, K. (eds), International handbook of labour market policy and evaluation, Edwar
Elgar, Cheltenham, UK, pp. 489–519.
319
The job creation potential of the service sector in Europe
Meager, N. and Bates, P. (1997), ‘Transitional labour markets: ‘‘bridges’’ between dependant
and self-employment’, paper for the Conference on New Institutional Arrrangements in the
Labour Market, Berlin, 11 and 12 April.
Menger, P. M. (1989), ‘Rationalité et incertitude de la vie d’artiste’, L’année sociologique, 39,
pp. 111–151.
Menger, P.-M. and Gurgand, M. (1996), ‘Work and compensated unemployment in the
performing arts — Exogenous and endogenous uncertainty in artistic labour markets’, in
Ginsburgh, V. A. and Menger, P.-M. (eds.), Economics of arts — Selected essays, Elsevier,
Amsterdam et al., pp. 347–381.
Menger, P. M. (1997), La profession de comédien, La documentation française, Paris.
Meyer, T. (1997), ‘Wieder ‘‘Selbstbedienungsökonomie’’ und ‘‘Brotverdienermodell’’?’, in
Behning, U. (ed.), Das Private ist ökonomisch, Edition Sigma, Berlin, pp. 189–205.
Meyer, T., Schulze, E., Böhm, U. and Fischer, B. (1999), ‘Anforderungen der Privathaushalte
an Dienstleistungen’, in Bullinger, H.-J. (ed.), Dienstleistungen — Innovation für Wachstum
und Beschäftigung, Gabler, Wiesbaden, pp. 224–242.
Millwarrd, N. and Woodland, S. (1995), ‘Gender segregation and male/female wage
differences’, in Humphries, J. and Rubery, J. (eds), The economics of equal opportunities,
Equal Opportunities Commission, Manchester.
Ministère de la culture (1995), ‘L’emploi des intermittents de l’audiovisuel et des spectacles’,
Développement culturel, 108, October.
Mintzberg, H. (1979), The structuring of organisations, Prentice Hall, Englewood Cliffs.
Monnoyer, M. and Philippe, J. (1991), ‘Localisation factors and development strategies of
producer services’, in Daniels, P. W. and Moulaert, F. (ed.), The changing geography of
advanced producer services, Belhaven Press, London and New York.
Morin, M. L. (1998), Rapport sur les prestations de travail dans les services, DARES,
Ministère de l’emploi, Paris.
Morin, M. L. (ed.) with Dupuy, Y., Larré, F. and Sublet, S. (1999), Prestation de travail et
activité de service, La documentation française.
Morvan, Y. (1991),‘Fondements d’économie industrielle’, Economica.
Mosley, H. (1994), ‘Employment protection and labour force adjustment in EC countries’, in
Schmid, G. (ed.), Labour market institutions in Europe, Armonk, New York and London,
p. 59.
320
Bibliography
Mosley, H. G. and Schütz, H. (2000), ‘Modernisation of the German Public Employment
Service: implementation of active policies in the German regions’, in De Koning, J. and
Mosley, H. G. (eds). Active labour market policy and labour market transitions, Edward
Elgar, Cheltenham, forthcoming.
Moss Kanter, R. (1995), ‘Nice work if you can get it: the software industry as a model for
tomorrow’s jobs’, The American Prospect, 23, pp. 52–59.
Müller-Siebrands, B. (1999), ‘Entwicklung eines sozialversicherungspflichtigen
Beschäftigungssektors ‘‘Dienstleistungen für private Haushalte’’ und die Hilfen der
Arbeitsämter’, in Zukunft im Zentrum (ed.), Innovative Dienstleistungen für private
Haushalte, Berlin, pp. 53–56.
Mundorf, N. (1999), ‘Information technologies in the home’, in Bullinger, H.-J. (ed.),
Dienstleistungen — Innovation für Wachstum und Beschäftigung, Gabler, Wiesbaden,
pp. 210–223.
Munz, S. (1996), ‘Beschäftigungspotentiale im Bereich Privater Haushalte’, IFO
Schnelldienst, 49, No 17/18, pp. 38–45.
Murphy, K. M. and Welch, F. (1992), ‘The structure of wages’, The Quarterly Journal of
Economics, pp. 285–326.
NBER (1961), ‘Output, input, and productivity measurement’, Studies in Income and Wealth,
Vol. 25, the Conference on Research in Income and Wealth, Princeton University Press,
Princeton.
Nelson, R. R. and Winter, S. G. (1982), An evolutionary theory of economic change, Belknap
Press, Cambridge.
Nickell, S. and Bell, B. (1995), ‘The collapse in demand for the unskilled and unemployment
across OECD’, Oxford Review of Economic Policy, Vol. 11(1), pp. 40–62.
Nicolas, B. (1995), ‘La flexibilité de l’organisation productive au cinéma’, in Benhamou, F.
et al., Approches comparatives en économie de la culture, Ministère de la culture, Paris, pp.
471–484.
Notarangelo, M. (1999a), ‘Unbalanced growth: a case of structural dynamics’, Structural
Change and Economic Dynamics, 10, pp. 209–223.
Notarangelo, M. (1999b), ‘Interaction between employment and effective demand: the effects
of technological change’, paper presented at the XIVth National Conference of Labour
Economics, AIEL, Milan Bicocca University, Milan, 7 and 8 October.
Oakley, A. (1974), Housewife, Penguin, London.
321
The job creation potential of the service sector in Europe
Oberheitmann, A. and Wenke, M. (1994), ‘Strukturveränderungen des westdeutschen Privaten
Verbrauchs — Methoden und ausgewählte Ergebnisse einer Neuberechnung der RWIKonsumverflechtungstabellen’ (‘Structural change of private consumption in West Germany
— Methods and selected results of a new computation of the RWI consumption
interpenetration tables’), RWI-Mitteilungen, No 2, pp. 103–126.
O’Brien, J. and Feist, A. (1995), Employment in the arts and cultural industries: an analysis
of the 1991 census, the Arts Council of England.
O’Brien, J. and Feist, A. (1997), Employment in the arts and cultural industries: an analysis
of the labour force survey and other sources, the Arts Council of England.
Observatoire de l’emploi culturel (1993), Les professions culturelles et les salariés des
activités culturelles d’après le recensement de la population, Ministère de la culture, Paris.
Observatoire de l’emploi culturel (1994), L’emploi dans le spectacle vivant et l’audiovisuel
d’après les données du GRISS, Ministère de la culture, Paris.
Observatoire de l’emploi culturel (1996), L’emploi culturel en 1995 d’après l’enquête sur
l’emploi de l’INSEE, Ministère de la culture, Paris.
Ochs, C. (1999), ‘Die unendliche Geschichte. zur Reform der geringfügigen Beschäftigung’,
WSI Mitteilungen, 52, April, pp. 223–233.
Odierna, S. (1995), ‘Private Haushalte als Arbeitgeber oder: die Rückkehr der Dienstmädchen
durch die Hintertür’, in Gräbe, S. (ed.), Private Haushalte und neue Arbeitsmodelle, Campus,
Frankfurt am Main/New York, pp. 87–100.
OECD (1992), The jobs study. See also European Commission (1998), Managing change,
final report of the High Level Group on Economic and Social Implications of Industrial
Change, November.
OECD (1994), Women and structural change: new perspectives, OECD, Paris.
OECD (1994), The OECD jobs study, OECD, Paris.
OECD (1996a), Employment and growth in the knowledge-based economy, OECD
documents.
OECD (1996b), Services: measuring real annual value added, OECD, Paris.
OECD (1996c), Historical statistics, OECD, Paris.
OECD (1999), ‘OECD science, technology and industry scoreboard’.
322
Bibliography
O’Farell, P. (1995), ‘Manufacturing demand for business services’, Cambridge Journal of
Economics, Vol. 19, pp. 523–543.
Okun, A. M. (1981), Prices and quantities, Basil Blackwell, Oxford.
Olsson, B. (1999), ‘Having or loving or being: working tie models and what it means for the
standard of living’, paper presented to the seventh meeting of the International Symposium on
Working Time, Gelsenkirchen, Germany, February.
Ott, N. (1998), ‘Der familienökonomische Ansatz von Gary S. Becker’, in Pies, I. and
Leschke, M. (eds), Gary Beckers ökonomischer Imperialismus, Mohr Siebeck, Tübingen,
pp. 63–90.
O’Reilly, J. (1996), ‘Labour adjustments through part-time work’, in Schmid, G., O’Reilly, J.
and Schöman, K. (eds), International handbook of labour market policy and evaluation,
Edwar Elgar, Cheltenham, UK, pp. 566–593.
O’Reilly, J. and Fagan, C. (eds) (1998), Part-time prospects: an international comparison of
part-time work in Europe, North America and the Pacific Rim, Routledge, London.
Pabst, S. (1999), ‘Mehr Arbeitsplätze für Geringqualifizierte nach Einführung der
Pflegeversicherung?’, WSI Mitteilungen, 52, April, pp. 234–240.
Pålsson, A,-M. (1997), Taxation and the market for domestic services’, in Persson, I. and
Jonung, C. (eds), Economics of the family and family politics, Routledge.
Paradeise, C. (1998), Les comédiens — Profession et marché du travail, PUF, Paris.
Pasinetti, L. (1981), Structural change and economic growth: a theoretical essay on the
dynamics of the wealth of nations, Cambridge University Press, Cambridge.
Peacock, A. and Weir, R. (1975), The composer in the market place, Faber, London.
Penrose, E. T. (1959), The theory of the growth of the firm, fourth edition, Basil Blackwell,
Oxford, 1968.
Perleman, S. and Pestieau, P. (1994), ‘A comparative study of postal services: a productivity
efficiency approach’, Annales d’Économie et de Statistique, No 33, pp. 187–202.
Persson, I. and Jonung, C. (1997), Economics of the family and family policies, Routledge,
London.
Persson, I. and Jonung, C. (1998), Women’s work and wage, Routledge, London.
Pestieau, P. (1993), ‘Performance and competition in services’, in ‘Market services and
European integration’, European Economy, No 3, pp. 125–148.
323
The job creation potential of the service sector in Europe
Philpott, J. (1999), ‘Temporary jobs, constant opportunity?’, Employment Policy Institute
Economic Report, Vol. 14, No 3, July.
Piketty, T. (1997), ‘Les créations d’emploi en France et aux Etats-Unis, ‘‘services de
proximité’’ contre ‘‘petits boulots’’ ’, Notes de la Fondation Saint-Simon, No 93, December.
Piketty, T. (1998), ‘L’emploi dans les services en France et aux Etats-Unis: une analyse
structurelle sur longue période’, Economie et Statistique, No 318, p. 8.
Pilat, D. (1996) ‘Labour productivity levels in OECD countries: estimates for manufacturing
and selected service sectors’, Economics Working Papers No 169, OECD, Paris.
Pisano, G. P. (1990), ‘The R & D boundaries of the firm: an empirical analysis’,
Administrative Science Quarterly, Vol. 35, March, pp. 153–173.
Plantenga, J. and Sloep, M. (1995), Women and the European employment rate: participation
and non-participation in the Netherlands — The causes and consequences of variations in
female activity and employment patterns, report to the European Commission, EC network on
the situation of women in the labour market, Manchester School of Management, UMIST.
Pryor, F. L. and Schaffer, D. L. (1999), Who’s not working and why: employment, cognitive
skills, wages, and the changing US labour market, Cambridge University Press, Cambridge,
UK.
Pyke, F. and Sengenberger, W. (eds) (1992), Industrial districts and local economic
regeneration, ILO, Geneva.
Quah, D. (1999), ‘The weightless economy in growth’, The Business Economist, Vol. 30,
No 1, 1999.
Reboud, L. (1997), La relation de service au cœur de l’analyse économique — Editions
l’Harmattan.
Reissert, B. and Schmid, G. (1999), ‘Berlin — Modellstadt der Arbeitsmarktpolitik?’ (‘Berlin
— A model city for labour market policy?’), in Momper, W., Kromphardt, J., Dybe, G. and
Steinke, R. (eds); Berlins Zweite Zukunft — Aufbruch in das 21. Jahrhundert, Edition Sigma,
Berlin, pp. 157–188.
Phelps, E. S. (1997), Rewarding work, Harvard University Press, Cambridge, Mass., and
London.
Reissert, B., Schmid, G. and Jahn, S. (1989), ‘Mehr Arbeitsplätze durch Dienstleistungen? Ein
Vergleich der Beschäftigungsentwicklung in den Ballungsregionen der Bundesrepublik
Deutschland’, Discussion Paper FS I 89-14, Wissenschaftszentrum Berlin für
Sozialforschung.
324
Bibliography
Richardson, G. B. (1972), ‘The organisation of industry’, Economic Journal, No 82,
pp. 883–896.
Robinson, P. (1997), ‘Water under the bridge: changes in employment in Britain and the
OECD’, LSE Centre for Economic Performance Discussion Paper No 325.
Roeger, W. and Wijkander, H. (1999a), ‘Unemployment in Europe: swimming against the tide
of skill-biased technical progress without relative wage adjustment’, unpublished paper.
Roeger, W. and Wijkander, H. (1999b), ‘Skill-biased technical change, sectoral heterogeneity
and wage setting: unemployment or wage inequality?’, unpublished paper.
Rogowski, R. and Schöman, K. (1996), ‘Legal regulation and flexibility of employment
contracts’, in Schmid, G., O’Reilly, J. and Schöman, K. (eds), International handbook of
labour market policy and evaluation, Edwar Elgar, Cheltenham, UK, pp. 623–651.
Ronneberger, K. (1992), ‘Von High-Tech-Regionen lernen?’ (‘Can we learn from high-tech
regions?’, Jahrbuch sozialwissenschaftliche Technikberichterstattung 1995, loc. cit.,
pp. 19–78.
Romer, P. (1986), ‘Increasing returns and long-run growth’, Journal of Political Eonomy,
94(5), pp. 1002–1037.
Romer, P. (1990), ‘Endogenous technical change’, Journal of Political Economy, 98(2),
October.
Rosen, S. (1981), ‘The economics of superstars’, American Economic Review, 75,
pp. 845–858.
Rosen, S. (1986), ‘The theory of equalising differences’, in Ashenfelter, O. and Layard, R.
(eds), Handbook of labor economics, Elsevier Science Publisher, BV, pp. 675–692.
Rostov, W. (1953), The process of economic growth, Oxford University Press, Oxford.
Stigler, G. (1951), ‘The division of labour is limited by the extent of the market’, Journal of
Political Economy, 59, June, pp. 185–193.
Rubery, J. and Fagan, C. (1998), Equal opportunities and employment in the European Union,
Federal Ministry of Labour, Health and Social Affairs, Vienna.
Rubery, J. and Fagan, C. (1993), ‘Occupational segregation of women and men in the
European Community’, Social Europe, Supplement 3/93, Office for Official Publications of
the European Communities, Luxembourg.
Rubery, J., Smith, M., Fagan, C. and Grimshaw, D. (1998), Women and European
employment, Routledge, London.
325
The job creation potential of the service sector in Europe
Rubery, J., Smith, M., and Fagan, C. (1999), Women’s employment in Europe: trends and
prospects, Routledge, London.
Sainsbury, D. (1996), Gender, equality and welfare states, Cambridge University Press,
Cambridge.
Sapir, A. (1991), ‘The structure of services in Europe: a conceptual framework’, Paper Series
No 498, Centre for Economic Policy Research Discussion, London.
Sassen, S. (1991), ‘The global city: New York, London, Tokyo’, Princeton University, NJ.
Sassen, S. (1993), ‘Global city: Internationale Verflechtungen und ihre innerstädtischen
Effekte’ (‘Global city: international interpenetration and its effects on the inner-city’), in
Häußermann, H. and Siebel, W. (eds), New York..., loc. cit., pp. 71–90.
Sauer, D. and Döhl, V. (1994), ‘Arbeit an der Kette — Systemische Rationalisierung
unternehmensübergreifender Produktion’ (‘Labour chains — Systemic rationalisation of
extra-plant production’), Soziale Welt, No 2, pp. 197–225.
Scharpf, F. W. (1990), ‘Structures of post-industrial society or does mass unemployment
disappear in the service and information economy?’, in Appelbaum, E. and Schettkat, R.
(eds.), Labor market adjustments to structural change and technological progress, Praeger,
New York, pp. 17–35.
Scharpf, F. W. (1993), ‘Von der Finanzierung der Arbeitslosigkeit zur Subventionierung
niedriger Erwerbseinkommen’, in Gewerkschaftliche Monatshefte, 44, July, pp. 433–443.
Scharpf, F. W. (1994), ‘Die negative Einkommenssteuer — Ein beschäftigungspolitisches
Instrument’, Wirtschaftsdienst, 74, No III, pp. 111–114.
Scharpf, F. W. (1998), ‘Die Zukunft der Arbeit im renovierten, Modell Deutschland’,
Gewerkschaftliche Monatshefte, 49, June/July, pp. 447–451.
Schedl, H. and Vogler-Ludwig, K. (1987), ‘Strukturverlagerungen zwischen sekundärem und
tertiärem Sektor — Zur Rolle der Dienstleistungen in der arbeitsteiligen Gesellschaft’
(‘Structural shifts between the secondary and tertiary sectors — On the role of services in the
labour-division society’), IFO Studien zur Strukturforschung, Vol. 8, IFO, Munich.
Schettkat, R. (1996), ‘Das Beschäftigungsproblem der Industriegesellschaften’, Aus Politik
und Zeitgeschichte, Vol. 26, pp. 25–35.
Schimmelpfennig, A. (1999), ‘Beschäftigung im Strukturwandel’ (‘Structural change in
employment’), in Bullinger, H.-J. (ed.), Dienstleistungen..., loc. cit., pp. 364–369.
326
Bibliography
Schmid, G. (1992), ‘Flexible Koordination: Instrumentarium erfolgreicher
Beschäftigungspolitik aus internationaler Perspektive’, Mitteilungen aus der Arbeitsmarktund Berufsforschung, 25, No 3, pp. 232–251.
Schmid, G. (1993), ‘Die beschäftigungspolitische Rolle des Staates für Frauen’, in Fricke, W.
(ed.), Jahrbuch für Arbeit und Technik 1993, Dietz, Bonn, pp. 52–64.
Schmid, G. (1998), ‘Transitional labour markets — A new European employment strategy’,
Discussion Paper FS I 98-206, Wissenschaftszentrum Berlin für Sozialforschung, 39 pp.
Schmid, G., Schütz, H. and Speckesser, S. (1999), ‘Broadening the scope of benchmarking:
radar charts and employment systems’, Labour. Review of Labour Economics and Industrial
Relations, Vol. 13, No 4, pp. 879–899.
Schmid, G. and Schömann, K. (1999), ‘Neue und alte Themenfelder für
Beschäftigungsbündnisse, in Piehl, E. and Timmann, H.-J. (eds), Der Europäische
Beschäftigungspakt, Nomos, Baden-Baden, pp. 51-62.
Schmid, G., Speckesser, S. and Hilbert, C. (2000), ‘Does active labour market policy matter?
An aggregate impact analysis of Germany’, in De Koning, J. and Mosley, H. G. (eds), Active
labour market policy and labour market transitions, Edward Elgar, Cheltenham, forthcoming.
Schreiner, O. (1997), ‘Arbeitsmarktpolitisches Konzept für gering qualifizierte Arbeit’, in
Friedrich Ebert Stiftung (ed.), Förderung der Beschäftigung von Geringqualifizierten in
Deutschland vor dem Hintergrund der Erfahrungen in Frankreich, den Niederlanden und
Schweden, Bonn, pp. 13–22.
Schupp, J., Volz, J., Wagner, G. and Zwiener, R. (1999), ‘Zuschüsse zu den
Sozialversicherungsbeiträgen im Niedriglohnbereich: Wenig zielgerichtet und teuer’, DIW
Wochenbericht, 66, No 27, pp. 499–509.
Schupp, J., Schwarze, J. and Wagner, G. (1997), ‘Erwerbsstatistik unterschätzt Beschäftigung
um 2 Millionen Personen’, DIW Wochenbericht, 64, No 38, pp. 689–694.
Schweitzer, R. V. (1995), ‘Strukturwandel in der Haushaltsproduktion’, in Gräbe, S. (ed.),
Private Haushalte und neue Arbeitsmodelle, Campus, Frankfurt am Main/New York,
pp. 39–49.
Scott, A. (ed.) (1994), Gender segregation and social change: men and women in changing
labour markets, Oxford University Press, Oxford.
Seewald, H. (1999), ‘Ergebnisse der Sozialhilfe- und Asylbewerberleistungsstatistik 1997’,
Wirtschaft und Statistik, No 2, pp. 96–110.
Semmler, W. (1998), ‘Arbeitsmarkt und Zukunftsperspektiven: USA und Deutschland’,
Leviathan, 4, pp. 557–566.
327
The job creation potential of the service sector in Europe
Sesselmeier, W., Klopfleisch, R. and Setzer, M. (1996), Mehr Beschäftigung durch eine
negative Einkommenssteuer, Lang, Frankfurt am Main.
Singelmann, J. (1978), From agriculture to services: the transformation of industrial
employment, Sage Publications, Beverly Hills.
Simmie, J. and Sennett, J. (1999), ‘Innovative clusters: global or local linkages?’, National
Institute Economic Review, No 170, pp. 87–98.
Seltzer, K. and Bentley, T. (1999), The creative age — Knowledge and skills for the new
economy, DEMOS.
Singer, L. (1981), ‘Supply decisions of professional artists’, American Economic Review,
pp. 341–346.
Slemrod, J. (1995), ‘What do cross-country studies teach about government involvement,
prosperity and economic growth?’, Brookings Papers on Economic Activity, No 2, 1995,
pp. 373–431.
Smith, M., Fagan, C. and Rubery, J. (1998), ‘Where and why is part-time work growing in
Europe?’, in O’Reilly, J. and Fagan, C. (eds), op. cit.
Snowdon, B. and Vane, H. (1999), ‘Interpreting macroeconomics’, an interview with Paul
Romer to appear in Conversations with Economists, edited by Brian Snowdon and Howard
Vane, Edward Elgar.
Snower, D. (1997), ‘The simple economics of benefit transfers’, in Snower, D. J. and de la
Dehesa, G. (eds), Unemployment policy, University Press, Cambridge, pp. 163–198.
Solow, R. (1956), ‘A contribution to the theory of economic growth’, Quarterly Journal of
Economics (70), pp. 65–94.
Solow, R. (1998), Work and welfare, Princeton University Press, Princeton (edited by Amy
Gutman, with comments by Himmelfarb, Lewis, Loury and Roemer).
Soziologisches Forschungsinstitut Göttingen (1999), Dienstleistungen als Chance:
Entwicklungspfade für die Beschäftigung (The opportunity of services: development paths for
employment), Abschlußbericht der PEM 13 — Kurzfassung — SOFI, Göttingen.
Spee, C. and Schmid, G. (1995), Strukturorientierte Arbeitsmarktpolitik für Berlin.
Beschäftigungsdynamik in Ballungsregionen, Berlin Schriftenreihe der Senatsverwaltung für
Arbeit und Frauen, Bd. 16, Teil 1 (this report has also been published as Discussion Paper FS
I 95–208, Wissenschaftszentrum Berlin für Sozialforschung).
Spermann, A. (1999), ‘Fighting long-term unemployment with targeted employment
subsidies’, Jahrbücher für Nationalökonomie und Statistik, Vol. 218/5+6, pp. 647–657.
328
Bibliography
Stahel, W. R. (1997), ‘Innovation braucht Nachhaltigkeit’, in Backhaus, K. and Bonus, H.
(eds), Die Beschleunigungsfalle oder der Triumph der Schildkröte, Stuttgart,
Schäffer-Poeschel Verlag, 2. erw. Auflage, pp. 67–92.
Stinchcombe, A. (1968), Constructing social theories, The University of Chicago Press,
Chicago.
Stigler, G. T. (1951), ‘The division of labour is limited by the extent of the market’, Journal
of Political Economy, Vol. 59, pp. 185–193.
Stigler, G. (1956), Trends in employment in the service sector, Princeton University Press,
Princeton.
Strambach, S. (1997), ‘Wissensintensive unternehmensorientierte Dienstleistungen — Ihre
Bedeutung für die Innovations- und Wettbewerbsfähigkeit Deutschlands’ (‘Knowledgeintensive business services — Their significance for Germany’s ability to innovate and
compete’), Vierteljahreshefte zur Wirtschaftsforschung, No 2, pp. 230–242.
Teece, D. J. (1988), ‘Technological change and the nature of the firm’, in Dosi, G., Freeman
C., Nelson, R., Silverberg, G. and Soete, L. (eds), Technical change and economic theory,
Pinter Publishers, pp. 256–281.
Teece, D. J., Rumelt, R., Dosi, G. and Winter, S. G. (1994), ‘Understanding corporate
coherence, theory and evidence’, Journal of Economic Behavior and Organization, Vol. 23,
No 1, January, pp. 1–30.
Temple, J. (1999), ‘The new growth evidence’, Journal of Economic Literature, Vol.
XXXVII, No 1.
Thèvenot, N. (1999), Les modes d’organisation des activités de service par les firmes — Une
approche théorique de leur hétérogénéité, thèse de Doctorat en sciences économiques,
Université de Paris I.
Throsby, D. (1994), ‘The production and consumption of the arts: a view of cultural
economics’, Journal of Economic Literature, Vol. 32, March, pp. 1–29.
Tiongson, E. R. (1997), ‘Baumol’s cost disease reconsidered’, Challenge, Vol. 40, No 6,
pp. 117–122.
Towse, R. (1992), ‘The labour market for artists’, Ricerche Economiche, XLVI, 1–2,
pp. 55–74.
Towse, R. (1993), Singers in the market place — The economics of the singing profession,
Charendon Press, Oxford.
329
The job creation potential of the service sector in Europe
Towse, R. (ed.) (1997), Baumol’s cost disease — The arts and other victims, Edward Elgar,
Cheltenham.
Tronti, L. and Cucchiarelli, A. (1992), ‘Inflazione, conflitto, concorrenza — La distribuzione
dei guadagni di produttività in Italia dal 1970 al 1991’, Economia & Lavoro, No 4.
Tschetter, J. (1987), ‘Producer services industries: Why are they growing so rapidly?’,
Monthly Labour Review, December, pp. 31–40.
Tuijnman, A. C. and Schömann, K. (1996), ‘Lifelong learning and skill formation’, in Schmid,
G., O’Reilly, J. and Schömann, K. (eds), International handbook of labour market policy and
evaluation, Edward Elgar, Cheltenham.
UK Department for Education and Employment and Cabinet Office (1996), The skills audit:
a report from an interdepartmental group, June.
UK Department of Trade and Industry (1998), Our competitive future: building the
knowledge-driven economy, Cm 4176 HMSO.
US Small Business Administration (1999), Internet (http://www.sba.gov) (Stand 5.7, 1999).
Vermeulen, H., Dex, S., Callan, T., Dankmeyer, B., Gustafsson, S., Lausten, M., Smith, N. and
Vlasblom, J. D. (1995), ‘Tax systems and married women’s labour force participation: a seven
country comparison’, Working Paper No 95-8 of the EC human capital and mobility female
labour force network, ESRC Research Centre on Micro-social Change, University of Essex,
UK.
Vierling, M. (1998), ‘Kombilohn: Die Frage ist nicht ‘‘ob’’, sondern ‘‘wie’’ ’,
Wirtschaftsdienst, 78, No IX, pp. 529–535.
Vogler, C. (1994), ‘Money in the household’, in Anderson, M., Bechhofer, F. and Gershuny, J.
(eds) (1994a), The social and political economy of the household, Oxford University Press,
Oxford.
Walwei, U. (1998), ‘Möglichkeiten und Grenzen der Schaffung eines Niedriglohnsektors’,
IAB Werkstattbericht, No 5.
Walwei, U. (1998), ‘Are part-time jobs better than no jobs?’, in O’Reilly, J. and Fagan, C.
(eds), op. cit.
Walwei, U. (1996), ‘Improving job-matching through placement service’, in Schmid, G.,
O’Reilly, J. and Schömann, K. (eds), International handbook of labour market policy and
evaluation, Edward Elgar, Cheltenham, pp. 402–430.
Warde, A. 1990. ‘Household work strategies and forms of labour: conceptual and empirical
issues’, Work, Employment and Society, 4(4), pp. 495–515.
330
Bibliography
Weinkopf, C. (1996), ‘Dienstleistungspools — Ein Ansatz zur Professionalisierung von
Dienstleistungen in Privathaushalten?’, WSI Mitteilungen, 49, January, pp. 36–43.
Weinkopf, C. (1997), ‘Beschäftigungsförderung im Bereich haushaltsbezogener
Dienstleistungen’, in Behning, U. (ed.), Das Private ist ökonomisch, Edition Sigma, Berlin,
pp. 133–151.
Weinkopf, C. (1999), ‘Privathaushalte und neue Dienste’, in Bullinger, H.-J. (ed.),
Dienstleistungen — Innovation für Wachstum und Beschäftigung, Gabler, Wiesbaden,
pp. 195–197.
Werner, H. (1997), ‘Die Arbeitsmarktentwicklung in den USA — Lehren für uns?’,
Mitteilungen aus der Arbeitsmarkt- und Berufsforschung, 30, No 3, pp. 585–600.
Wyckoff, A. (1996), ‘The growing strength of services’, OECD Observer, June/July,
pp. 11–15.
Westwood, A. (1999), Education and work, report prepared by the Employment Policy
Institute for the Edexel Foundation.
Williamson, O. E. (1975), Markets and hierarchies: analysis and antitrust implications, The
Free Press, New York.
Williamson, O. E. (1985), The economic institutions of capitalism, The Free Press, New York.
Williamson, O. E. (1991), ‘Comparative economic organisation: the analysis of discrete
structural alternatives’, Administrative Science Quarterly, Vol. 36, pp. 269–296.
Williamson, O. E. (1993), ‘Transaction cost economics and organisation theory’, Industrial
and Corporate Change, Vol. 2, No 3, pp. 107–156.
Wood, A. (1994), North-South trade, employment and inequality: changing fortunes in a skilldriven world, Oxford.
Woolf, M. and Holly, S. (1994), Employment patterns and training needs 1993/94, Skillset.
Xerfi (1997), ‘Secteurs 700-1997’, Série 701, Services aux entreprises, Vol. 2.
Zart, M. and Huege, P. (1999), ‘Auswirkungen der Globalisierung auf die Regionen der
Bundesrepublik Deutschland’ (‘Effects of globalisation on the regions of the Federal Republic
of Germany’), Informationen zur Raumentwicklung, No 1, pp. 1–7.
Zentrum für Europäische Wirtschaftsforschung (ZEW) (1998), ‘Beschäftigungseffekte von
Lohnsubventionen im Niedriglohnbereich — Gutachten im Auftrag der Hans-BöcklerStiftung’ (duplicated manuscript), Mannheim.
331
The job creation potential of the service sector in Europe
Zentrum für Europäische Wirtschaftsforschung (ZEW) (1999), ‘Dienstleistungen in der
Zukunft — Innovationsaktivitäten im Dienstleistungssektor: Befragung 1998’ (‘The future of
services — Innovation activities in the service sector: Questionnaire 1998’), ZEW,
Mannheim.
Zukunft im Zentrum (1999a), ‘Dienstleistungspools, -zentren und –agenturen’, in Zukunft im
Zentrum (ed.), Innovative Dienstleistungen für private Haushalte, Berlin, pp. 124–129.
Zukunft im Zentrum (1999b), Identifizierung und Systematisierung aktueller und künftiger
Dienstleistungspotentiale in Berliner Dienstleistungsagenturen, Berlin.
Zukunftskommission der Friedrich Ebert Stiftung (1998), Wirtschaftliche Leistungsfähigkeit,
sozialer Zusammenhalt, ökologische Nachhaltigkeit (Economic efficiency, social cohesion,
ecological sustainability), Dietz, Bonn.
332
European Commission
The job creation potential of the service sector in Europe: Final report 2000
Luxembourg: Office for Official Publications of the European Communities
2001 — 332 pp. — 21 x 29.7 cm
ISBN 92-894-0603-8